---
week: 2026-w46
start: 2026-11-09
end: 2026-11-15
notes: 553
html: https://newruntime.com/weeks/2026-w46/
json: https://newruntime.com/weeks/2026-w46.json
---

# Week in review: Nov 9–15, 2026

553 Field Notes, grouped by topic. Each links to its Markdown twin; replace .md with .json for structured fields.

## Claude (68)

- 2026-11-14 [An update on our recent security incident](https://newruntime.com/posts/an-update-on-our-recent-security-incident-d48bfb.md)
- 2026-11-14 [Analysts say Anthropic's GSA 40% discount is not yet an agency's all-in price](https://newruntime.com/posts/analysts-say-anthropic-s-gsa-40-discount-is-40f144.md)
- 2026-11-14 [DeepSeek Harness ships dsh-v0.2.2-rc.1, the first 0.2.2 release candidate](https://newruntime.com/posts/deepseek-harness-ships-dsh-v0-2-2-rc-0e252a.md)
- 2026-11-14 [GameBench 2 ranks Claude Fable 5.2 first; low-effort programs beat Fable 5.1 at lower cost](https://newruntime.com/posts/gamebench-2-ranks-claude-fable-5-2-first-1510a6.md)
- 2026-11-14 [Hedwigai test finds Particle-1 as accurate as DeepSeek V4.1 Flash but slower and slightly more token-heavy](https://newruntime.com/posts/hedwigai-test-finds-particle-1-as-accurate-as-8ba3d3.md)
- 2026-11-14 [Kimi K3.1 Pro leads Terminal-Bench 4.1 at 68%, ahead of Claude Sonnet 5.5 at 66%.](https://newruntime.com/posts/kimi-k3-1-pro-leads-terminal-bench-4-a1aa9b.md)
- 2026-11-14 [Kitesurf - stateless browser running entirely on Workers](https://newruntime.com/posts/kitesurf-stateless-browser-running-entirely-on-workers-bee538.md)
- 2026-11-14 [Plotly's analytics exam puts Muse Ember off the cost frontier versus GPT-6 Luna](https://newruntime.com/posts/plotly-s-analytics-exam-puts-muse-ember-off-67701f.md)
- 2026-11-14 [Ripple-65B Ollama tags list 2M context while Ricursive advertises 1M, and Terminal-Bench scores remain unreproduced](https://newruntime.com/posts/ripple-65b-ollama-tags-list-2m-context-while-3646ec.md)
- 2026-11-14 [Science: OpenAI still withheld the 1,147-page proof; Lean check is not full correctness](https://newruntime.com/posts/science-openai-still-withheld-the-1-147-page-3721ad.md)
- 2026-11-14 [Spacefast is a new kind of infrastructure for the vibe coding age](https://newruntime.com/posts/spacefast-is-a-new-kind-of-infrastructure-for-4d3f62.md)
- 2026-11-14 [Version 3.5.1 adds support for parallel sessions and multiple machines that share the same vault.](https://newruntime.com/posts/version-3-5-1-adds-support-for-parallel-6e35f3.md)
- 2026-11-14 [WorkflowArena: Executable Workflow Orchestration](https://newruntime.com/posts/workflowarena-executable-workflow-orchestration-05cbd4.md)
- 2026-11-13 [Building a self-service LLM gateway with Amazon Bedrock and LiteLLM](https://newruntime.com/posts/building-a-self-service-llm-gateway-with-amazon-791630.md)
- 2026-11-13 [How to train a Transformer (as a software engineer with no ML experience)](https://newruntime.com/posts/how-to-train-a-transformer-as-a-software-32d1f6.md)
- 2026-11-13 [Microsoft starts rolling out Claude Fable 5.2 in Copilot Cowork and Copilot Studio](https://newruntime.com/posts/microsoft-starts-rolling-out-claude-fable-5-2-559f16.md)
- 2026-11-13 [XYBench: Helping with the problem behind the question](https://newruntime.com/posts/xybench-helping-with-the-problem-behind-the-question-172bdf.md)
- 2026-11-12 [An AI Agent Injected Drugs Into a Live Mouse’s Brain](https://newruntime.com/posts/an-ai-agent-injected-drugs-into-a-live-899659.md)
- 2026-11-12 [Bring your business context to ChatGPT, Claude, Microsoft Copilot Cowork and Grok Bot](https://newruntime.com/posts/bring-your-business-context-to-chatgpt-claude-microsoft-e2b750.md)
- 2026-11-12 [cinecode: Films made with code](https://newruntime.com/posts/cinecode-films-made-with-code-c0c061.md)
- 2026-11-12 [Claude app for iOS is getting “Saved logins” section where users can see credentials that Claude can use to sign in on other websites.](https://newruntime.com/posts/claude-app-for-ios-is-getting-saved-logins-13fd99.md)
- 2026-11-12 [Claude Code 2.1.323 adds experimental Code Mode and plugin engine version ranges](https://newruntime.com/posts/claude-code-2-1-323-adds-experimental-code-af152c.md)
- 2026-11-12 [Claude for Manufacturing: AI plugin for plant teams](https://newruntime.com/posts/claude-for-manufacturing-ai-plugin-for-plant-teams-664799.md)
- 2026-11-12 [DevoxxGenie 1.15.0 adds favorite models, per-project scan settings, Markdown fetch_page, and Claude 5](https://newruntime.com/posts/devoxxgenie-1-15-0-adds-favorite-models-per-eeb6a1.md)
- 2026-11-12 [durable-harness 0.0.1 release adds Agent Runs tracing](https://newruntime.com/posts/durable-harness-0-0-1-release-adds-agent-713a6e.md)
- 2026-11-12 [Introducing Claude Meetings, now in beta](https://newruntime.com/posts/introducing-claude-meetings-now-in-beta-386ad2.md)
- 2026-11-12 [J.Crew’s Style Advisor: an AI stylist built with Claude](https://newruntime.com/posts/j-crew-s-style-advisor-an-ai-stylist-2e807a.md)
- 2026-11-12 [MiniMax publishes MiniMax-H3-Image, an open-weight 16-megapixel image generation and editing model](https://newruntime.com/posts/minimax-publishes-minimax-h3-image-an-open-weight-7378fa.md)
- 2026-11-12 [Perplexity published Escaping SPACE: Part II, analyzing how 12 models behaved during sandbox red-teaming.](https://newruntime.com/posts/perplexity-published-escaping-space-part-ii-analyzing-how-ea11c2.md)
- 2026-11-12 [Simular launches Box, a shared cloud Linux desktop for AI agents via one remote MCP server](https://newruntime.com/posts/simular-launches-box-a-shared-cloud-linux-desktop-7e876d.md)
- 2026-11-12 [WTF are evals? How to test agents and the software they use](https://newruntime.com/posts/wtf-are-evals-how-to-test-agents-and-ece8bb.md)
- 2026-11-11 [Alibaba opens Qwen4-Max preview on QwenCloud Token Plan until Nov 25](https://newruntime.com/posts/alibaba-opens-qwen4-max-preview-on-qwencloud-token-992fc3.md)
- 2026-11-11 [Amazon saves millions by replacing Sonnet 4.6 with Strands Decider](https://newruntime.com/posts/amazon-saves-millions-by-replacing-sonnet-4-6-956138.md)
- 2026-11-11 [Anthropic’s Claude Mythos 5.2 is an active model released on November 9, 2026](https://newruntime.com/posts/anthropic-s-claude-mythos-5-2-is-an-07a4f8.md)
- 2026-11-11 [Anthropic’s Supported Regions Policy, effective November 12, 2026, limits access to its apps, API](https://newruntime.com/posts/anthropic-s-supported-regions-policy-effective-november-12-fa0e7d.md)
- 2026-11-11 [Claude account shows a Fable 5.2 3D landmark built from 40 photos](https://newruntime.com/posts/claude-account-shows-a-fable-5-2-3d-db13ba.md)
- 2026-11-11 [Lazare, veni foras: Reprogramming Dormant Neurons to Backdoor AI Agents](https://newruntime.com/posts/lazare-veni-foras-reprogramming-dormant-neurons-to-backdoor-022e16.md)
- 2026-11-11 [Medium launches TK, an AI-powered private notebook and writing app](https://newruntime.com/posts/medium-launches-tk-an-ai-powered-private-notebook-0b1099.md)
- 2026-11-11 [Over $100M in credits offered to eligible companies to hunt critical vulnerabilities.](https://newruntime.com/posts/over-100m-in-credits-offered-to-eligible-companies-9f0c22.md)
- 2026-11-11 [Raindrop study: agent swarms rarely produce a true leader, and prompt order shapes who plans](https://newruntime.com/posts/raindrop-study-agent-swarms-rarely-produce-a-true-0f131a.md)
- 2026-11-11 [Updating the ECI methodology to use a Bayesian model](https://newruntime.com/posts/updating-the-eci-methodology-to-use-a-bayesian-6c0092.md)
- 2026-11-11 [Working with Claude Sheets, Whiteboards, and Interactive Connectors](https://newruntime.com/posts/working-with-claude-sheets-whiteboards-and-interactive-connectors-14c6eb.md)
- 2026-11-10 [Anthropic adds Fable 5.2 launch offers: extra Fable limit resets for paid Claude plans](https://newruntime.com/posts/anthropic-adds-fable-5-2-launch-offers-extra-dcccc1.md)
- 2026-11-10 [Claude can design electronic prototypes from start to finish](https://newruntime.com/posts/claude-can-design-electronic-prototypes-from-start-to-679cd0.md)
- 2026-11-10 [Claude Fable 5.2 is now available in Devin](https://newruntime.com/posts/claude-fable-5-2-is-now-available-in-0cc74d.md)
- 2026-11-10 [Claude Fable 5.2 is now available.](https://newruntime.com/posts/claude-fable-5-2-is-now-available-08970e.md)
- 2026-11-10 [Claude Fable 5.2 scored 174 on the Epoch Capabilities Index, beating GPT-6 Astra's prior record of 169.](https://newruntime.com/posts/claude-fable-5-2-scored-174-on-the-2701ba.md)
- 2026-11-10 [Claude Fable 5.2 takes #1 on the Vals Index at 71.31% accuracy](https://newruntime.com/posts/claude-fable-5-2-takes-1-on-the-67e6f8.md)
- 2026-11-10 [Introducing Claude Fable 5.2 on AWS](https://newruntime.com/posts/introducing-claude-fable-5-2-on-aws-4ac017.md)
- 2026-11-10 [llm-anthropic 0.31 adds Claude Fable 5.2 support to the llm command-line tool](https://newruntime.com/posts/llm-anthropic-0-31-adds-claude-fable-5-9ddd40.md)
- 2026-11-10 [Locating the source of a mysterious beep with Claude](https://newruntime.com/posts/locating-the-source-of-a-mysterious-beep-with-61e58e.md)
- 2026-11-10 [Luigi Petrucco reconstructs Meta’s Muse Spark Open Charter from chain-of-thought](https://newruntime.com/posts/luigi-petrucco-reconstructs-meta-s-muse-spark-open-7743a3.md)
- 2026-11-10 [Meta introduced Muse Ember 1.0, a reasoning model aimed at repeated agent work such as tool calls](https://newruntime.com/posts/meta-introduced-muse-ember-1-0-a-reasoning-1cdb9e.md)
- 2026-11-10 [Meta released Muse Ember, its smallest reasoning model, matching Claude Haiku 5.5 intelligence at one-third the cost.](https://newruntime.com/posts/meta-released-muse-ember-its-smallest-reasoning-model-98fb00.md)
- 2026-11-10 [OpenCode Go Teams: $20-per-member flat-rate model subscriptions](https://newruntime.com/posts/opencode-go-teams-20-per-member-flat-rate-21805c.md)
- 2026-11-10 [Ripple-65B Preview from Ricursive Intelligence lands on Ollama as ripple-preview](https://newruntime.com/posts/ripple-65b-preview-from-ricursive-intelligence-lands-on-7caa24.md)
- 2026-11-10 [Vals reports the clearest improvement over Fable 5.1 in coding.](https://newruntime.com/posts/vals-reports-the-clearest-improvement-over-fable-5-fd58c3.md)
- 2026-11-10 [With Conductor winding down, your threads need a new home.](https://newruntime.com/posts/with-conductor-winding-down-your-threads-need-a-97da46.md)
- 2026-11-09 [Claude Fable 5.2 is live on AI Gateway.](https://newruntime.com/posts/claude-fable-5-2-is-live-on-ai-55233b.md)
- 2026-11-09 [Clicking “Yes, and don’t ask again” in Claude Code can save your password in a file that ends up public on GitHub, where anyone can read it.](https://newruntime.com/posts/clicking-yes-and-don-t-ask-again-in-852ac3.md)
- 2026-11-09 [Cursor has added Claude Fable 5.2 to its model lineup.](https://newruntime.com/posts/cursor-has-added-claude-fable-5-2-to-c923e0.md)
- 2026-11-09 [Introducing Claude Fable 5.2 and Claude Mythos 5.2](https://newruntime.com/posts/introducing-claude-fable-5-2-and-claude-mythos-6b1c41.md)
- 2026-11-09 [Meta launched Muse Ember, its smallest model, built for agent subtasks and tool calls.](https://newruntime.com/posts/meta-launched-muse-ember-its-smallest-model-built-4baf24.md)
- 2026-11-09 [Muse Ember is now available on AI Gateway.](https://newruntime.com/posts/muse-ember-is-now-available-on-ai-gateway-3d3df9.md)
- 2026-11-09 [parse_number() as a SQLite extension: C vs Rust](https://newruntime.com/posts/parse-number-as-a-sqlite-extension-c-vs-378a34.md)
- 2026-11-09 [Project EVE: Claude plays MMORPGs](https://newruntime.com/posts/project-eve-claude-plays-mmorpgs-430334.md)
- 2026-11-09 [reelmimic: Show it a video you love](https://newruntime.com/posts/reelmimic-show-it-a-video-you-love-f0d2b4.md)
- 2026-11-09 [The model checked its own plan, writing and running code to confirm, atom by atom, that each compound followed from its starting materials.](https://newruntime.com/posts/the-model-checked-its-own-plan-writing-and-c98cff.md)

## Models (68)

- 2026-11-15 [Perplexity released Decider v1.3 (27B) and Flash v1.3 (4B), open-weight decision models under Apache 2.0.](https://newruntime.com/posts/perplexity-released-decider-v1-3-27b-and-flash-28f495.md)
- 2026-11-14 [A unifying view of contrastive learning, importance sampling, and bridge sampling for energy-based models](https://newruntime.com/posts/a-unifying-view-of-contrastive-learning-importance-sampling-430eed.md)
- 2026-11-14 [Action-Guided Reachability for Deterministic World Models of Vision-Based Systems](https://newruntime.com/posts/action-guided-reachability-for-deterministic-world-models-of-04eb8f.md)
- 2026-11-14 [AIPOCH Open-Science v1.2.2 adds strict Notebook networking and Qwen4 Max Preview](https://newruntime.com/posts/aipoch-open-science-v1-2-2-adds-strict-34c977.md)
- 2026-11-14 [AWS adds bedrock-runtime access for Daybreak Red and Daybreak Blue](https://newruntime.com/posts/aws-adds-bedrock-runtime-access-for-daybreak-red-30813e.md)
- 2026-11-14 [CleanYourTV 3.5: eight new themes, 472 TV models and a lighter Home](https://newruntime.com/posts/cleanyourtv-3-5-eight-new-themes-472-tv-8ba442.md)
- 2026-11-14 [fx v0.0.17 Subagents & Token Pass, first-class.](https://newruntime.com/posts/fx-v0-0-17-subagents-token-pass-first-626591.md)
- 2026-11-14 [higurashi releases Haru-4B, an open full-duplex spoken dialogue base model](https://newruntime.com/posts/higurashi-releases-haru-4b-an-open-full-duplex-02f6d1.md)
- 2026-11-14 [Hugging Face tokenizers v1.0.0-rc.4 restores Python APIs and splits Tokenizer from CompiledTokenizer](https://newruntime.com/posts/hugging-face-tokenizers-v1-0-0-rc-4-f5dd41.md)
- 2026-11-14 [InferSSD streams MoE expert weights from NVMe, running 700B+ models on consumer hardware.](https://newruntime.com/posts/inferssd-streams-moe-expert-weights-from-nvme-running-c0923f.md)
- 2026-11-14 [Researchers at IIT Madras and UIUC have proposed FedBNN, a framework that trains binary neural networks within the federated learning loop.](https://newruntime.com/posts/researchers-at-iit-madras-and-uiuc-have-proposed-d797e1.md)
- 2026-11-14 [Restrict models per AI Gateway key](https://newruntime.com/posts/restrict-models-per-ai-gateway-key-2cf7e6.md)
- 2026-11-14 [TechCrunch reports Chinese frontier models show fingerprints of export-banned Nvidia Blackwell chips](https://newruntime.com/posts/techcrunch-reports-chinese-frontier-models-show-fingerprints-of-fb9527.md)
- 2026-11-14 [UC Berkeley and NVIDIA paper wins CoRL 2026 Agentic Robotics workshop best paper](https://newruntime.com/posts/uc-berkeley-and-nvidia-paper-wins-corl-2026-d54a22.md)
- 2026-11-13 [Atlas: World Labs](https://newruntime.com/posts/atlas-world-labs-8d2578.md)
- 2026-11-13 [Cmpnd releases dspy-wsp, a DSPy optimizer that lets a frontier model edit a small model's weights](https://newruntime.com/posts/cmpnd-releases-dspy-wsp-a-dspy-optimizer-that-745605.md)
- 2026-11-13 [Continuous Intelligence releases ci-proto, five open time-series foundation models up to 167M parameters](https://newruntime.com/posts/continuous-intelligence-releases-ci-proto-five-open-time-c1a027.md)
- 2026-11-13 [DeepSeek brings full App capabilities to deepseek-flash API: search, files, URLs, safety.](https://newruntime.com/posts/deepseek-brings-full-app-capabilities-to-deepseek-flash-8835c5.md)
- 2026-11-13 [Deploy frontier AI models on AWS Trainium with SGLang-Neuron](https://newruntime.com/posts/deploy-frontier-ai-models-on-aws-trainium-with-2a6551.md)
- 2026-11-13 [Edge0 open-sources Edge0-145B and Edge0-120B latent-MoE models and the edge0-strata streaming inference runtime](https://newruntime.com/posts/edge0-open-sources-edge0-145b-and-edge0-120b-f895dd.md)
- 2026-11-13 [Encevra releases Athena-27B-Preview, a LoRA adapter for Qwen3.8-27B that answers from documents with verbatim quotes and abstains](https://newruntime.com/posts/encevra-releases-athena-27b-preview-a-lora-adapter-ce3c6c.md)
- 2026-11-13 [GATE-AD is a new few-shot industrial defect detector accepted at NeurIPS 2026 ( paper ).](https://newruntime.com/posts/gate-ad-is-a-new-few-shot-industrial-b5fc03.md)
- 2026-11-13 [Liquid AI released LiquidCodec, a ~74M-param streaming audio codec for LFM3-Audio-2B and LFM3-Omni-2B.](https://newruntime.com/posts/liquid-ai-released-liquidcodec-a-74m-param-streaming-40a5e4.md)
- 2026-11-13 [Mistral hosts DeepSeek V4.1 Flash 1017 in public preview and GLM 5.3 Flash as generally available](https://newruntime.com/posts/mistral-hosts-deepseek-v4-1-flash-1017-in-22fee7.md)
- 2026-11-13 [New Live 3 full-duplex model reacts in 186 ms for short cues like "mm-hmm."](https://newruntime.com/posts/new-live-3-full-duplex-model-reacts-in-7341b8.md)
- 2026-11-13 [Perplexity Decider v1.3: Training, eval, and use cases](https://newruntime.com/posts/perplexity-decider-v1-3-training-eval-and-use-816024.md)
- 2026-11-13 [Recursive model card reports Flint’s 89.7% SWE-bench Pro v2 score](https://newruntime.com/posts/recursive-model-card-reports-flint-s-89-7-cdd01f.md)
- 2026-11-13 [Salesforce releases SOV-Transcribe-1, an Apache 2.0 open-weight speech-to-text model](https://newruntime.com/posts/salesforce-releases-sov-transcribe-1-an-apache-2-c6a174.md)
- 2026-11-13 [Terradot, Microsoft build AI model to test crushed rock carbon removal](https://newruntime.com/posts/terradot-microsoft-build-ai-model-to-test-crushed-416e9f.md)
- 2026-11-13 [The full DeepSeek App experience, now via API](https://newruntime.com/posts/the-full-deepseek-app-experience-now-via-api-3914ad.md)
- 2026-11-12 [Code and dataset released at anonymous.4open.science; paper on arXiv.](https://newruntime.com/posts/code-and-dataset-released-at-anonymous-4open-science-41b622.md)
- 2026-11-12 [feat(opencode): bring back MiniMax M3.1 Flash Free](https://newruntime.com/posts/feat-opencode-bring-back-minimax-m3-1-flash-a62a25.md)
- 2026-11-12 [Huawei Computing Systems Lab releases MACaQ, an Apache-2.0 model-agnostic quantizer](https://newruntime.com/posts/huawei-computing-systems-lab-releases-macaq-an-apache-68407c.md)
- 2026-11-12 [LLMON: An LLM-Native Markup Language, now in Granite 4.1 3B and 8B](https://newruntime.com/posts/llmon-an-llm-native-markup-language-now-in-448050.md)
- 2026-11-12 [LLMs can learn some multi-step reasoning on filler tokens instead of visible CoT](https://newruntime.com/posts/llms-can-learn-some-multi-step-reasoning-on-e5a055.md)
- 2026-11-12 [Nvidia investing $3 billion in Meta in Muse partnership](https://newruntime.com/posts/nvidia-investing-3-billion-in-meta-in-muse-56f888.md)
- 2026-11-12 [Phonon 0.5.0 loads Kyutai Pocket TTS small-model architectures and fixes Mimi decoder attention](https://newruntime.com/posts/phonon-0-5-0-loads-kyutai-pocket-tts-147e63.md)
- 2026-11-12 [Shipping DoorDash’s First Production Transformer for Homepage Store Ranking Deep Learning Engineering](https://newruntime.com/posts/shipping-doordash-s-first-production-transformer-for-homepage-f97659.md)
- 2026-11-12 [Shopify’s generative discovery model](https://newruntime.com/posts/shopify-s-generative-discovery-model-fd2f6f.md)
- 2026-11-11 [ConvictionBench tests whether LLMs change answers only for good reasons, including forged vs tool-authenticated records](https://newruntime.com/posts/convictionbench-tests-whether-llms-change-answers-only-for-77377b.md)
- 2026-11-11 [Good old-fashioned engineering can close the 100,000-year “data gap” in robotics](https://newruntime.com/posts/good-old-fashioned-engineering-can-close-the-100-1013f1.md)
- 2026-11-11 [How to make flash models even faster: optimizing DeepSeek V4.1 Flash 1017](https://newruntime.com/posts/how-to-make-flash-models-even-faster-optimizing-131ad4.md)
- 2026-11-11 [Introducing Kimi K3.1 Pro on Amazon Bedrock](https://newruntime.com/posts/introducing-kimi-k3-1-pro-on-amazon-bedrock-4f1eb4.md)
- 2026-11-11 [MBZUAI's Institute of Foundation Models releases PAN 2, an open 29B world model that simulates action outcomes as video](https://newruntime.com/posts/mbzuai-s-institute-of-foundation-models-releases-pan-26018c.md)
- 2026-11-11 [Neurons as an Interpretable Basis](https://newruntime.com/posts/neurons-as-an-interpretable-basis-ffb4bb.md)
- 2026-11-11 [Trained via supervised cold start plus AdaGRPO reinforcement learning with mode-relative advantage.](https://newruntime.com/posts/trained-via-supervised-cold-start-plus-adagrpo-reinforcement-5f960f.md)
- 2026-11-10 [Ant Group's Robbyant open-sources LingBot-Touch tactile foundation model for robot hands](https://newruntime.com/posts/ant-group-s-robbyant-open-sources-lingbot-touch-bd9e36.md)
- 2026-11-10 [BALD-SAM adds Bayesian active learning on top of frozen SAM to pick the most informative next prompt.](https://newruntime.com/posts/bald-sam-adds-bayesian-active-learning-on-top-b32aaf.md)
- 2026-11-10 [Decision Models in Hugging Face](https://newruntime.com/posts/decision-models-in-hugging-face-f020e5.md)
- 2026-11-10 [From Full to Hybrid Attention in LLMs](https://newruntime.com/posts/from-full-to-hybrid-attention-in-llms-795700.md)
- 2026-11-10 [Mitra-v2: Scaling synthetic pretraining for tabular foundation models](https://newruntime.com/posts/mitra-v2-scaling-synthetic-pretraining-for-tabular-foundation-63e6cb.md)
- 2026-11-10 [Nathan Barry reports that a linear program can reproduce much of GPT-6 Astra’s edit to Qwen3.5-0.8B](https://newruntime.com/posts/nathan-barry-reports-that-a-linear-program-can-6a8a22.md)
- 2026-11-10 [New GLM-5.3-SLIM-GGUF fits a GLM-5.3-class model on one NVIDIA B200 at 179.9 GB.](https://newruntime.com/posts/new-glm-5-3-slim-gguf-fits-a-0507b5.md)
- 2026-11-10 [NVIDIA Introduces Hardware-Aware LLM Compression](https://newruntime.com/posts/nvidia-introduces-hardware-aware-llm-compression-67648c.md)
- 2026-11-10 [NVIDIA Nemotron 3.5 Super VL is now available in Amazon SageMaker JumpStart](https://newruntime.com/posts/nvidia-nemotron-3-5-super-vl-is-now-284b5a.md)
- 2026-11-10 [Pinecone integrated embedding now supports OpenAI's text-embedding-3 models](https://newruntime.com/posts/pinecone-integrated-embedding-now-supports-openai-s-text-830072.md)
- 2026-11-10 [Zyphra open-sources ZONOS3-Mini, a 100M-active-parameter on-device text-to-speech and voice-cloning model](https://newruntime.com/posts/zyphra-open-sources-zonos3-mini-a-100m-active-6c2984.md)
- 2026-11-09 [169Pi has released Alpie Seed 9B](https://newruntime.com/posts/169pi-has-released-alpie-seed-9b-c1bb5a.md)
- 2026-11-09 [Cosmos3-Nano is a 16B omnimodal world model for Physical AI applications.](https://newruntime.com/posts/cosmos3-nano-is-a-16b-omnimodal-world-model-9731b7.md)
- 2026-11-09 [GGUF port of Tencent's DSpine draft model for Qwen3.8-27B is live on Hugging Face.](https://newruntime.com/posts/gguf-port-of-tencent-s-dspine-draft-model-37507f.md)
- 2026-11-09 [NVIDIA and AI4Bharat publish Nemotron-Pretraining-Indic-v1 and Indic-LLM-Eval-Suite](https://newruntime.com/posts/nvidia-and-ai4bharat-publish-nemotron-pretraining-indic-v1-77b559.md)
- 2026-11-09 [NVIDIA introduces GR00T 2 world action model and Isaac Lab 3.0 at CoRL](https://newruntime.com/posts/nvidia-introduces-gr00t-2-world-action-model-and-da1a8f.md)
- 2026-11-09 [Parakeet Ultra now available as GGUF quantizations for transcribe.cpp, from F32 (2.51 GB) to Q4_K_M (486 MB).](https://newruntime.com/posts/parakeet-ultra-now-available-as-gguf-quantizations-for-952aab.md)
- 2026-11-09 [PS5LM-Plus: llama.cpp on a jailbroken PS5: run any GGUF model natively on the console, starting with Qwen 3.8](https://newruntime.com/posts/ps5lm-plus-llama-cpp-on-a-jailbroken-ps5-0f3192.md)
- 2026-11-09 [Snowflake runs NVIDIA Kumo Relational and Kumo Engine in-account, naming Affirm, NOV, VideoAmp and Yieldmo](https://newruntime.com/posts/snowflake-runs-nvidia-kumo-relational-and-kumo-engine-49f0e8.md)
- 2026-11-09 [T-Tech lab announces F-GRPO and ESSA to cut the cost of LLM fine-tuning](https://newruntime.com/posts/t-tech-lab-announces-f-grpo-and-essa-aaaf91.md)
- 2026-11-09 [Upstream Ultra cuts FLEURS 25-language WER from 11.62% to 9.55% vs base parakeet-tdt-0.6b-v3.](https://newruntime.com/posts/upstream-ultra-cuts-fleurs-25-language-wer-from-9eb762.md)
- 2026-11-09 [VulcanBench ranks Reflection Beam far below frontier models on Frontier v4 reconstruction](https://newruntime.com/posts/vulcanbench-ranks-reflection-beam-far-below-frontier-models-8ecbd2.md)

## Google (49)

- 2026-11-14 [Google Cloud Batch adds jobs on single-host TPU slices](https://newruntime.com/posts/google-cloud-batch-adds-jobs-on-single-host-4589d6.md)
- 2026-11-14 [Google Cloud publishes open AI-native stack on Kubernetes: OpenRL, llm-d 1.0, KubeRay on GKE](https://newruntime.com/posts/google-cloud-publishes-open-ai-native-stack-on-e79e26.md)
- 2026-11-14 [Google connects Wallet to Gemini for U.S. personal finance insights](https://newruntime.com/posts/google-connects-wallet-to-gemini-for-u-s-20fb16.md)
- 2026-11-14 [Google releases stable A2UI 1.0 with Express, swappable inference and named production users](https://newruntime.com/posts/google-releases-stable-a2ui-1-0-with-express-418339.md)
- 2026-11-14 [Google Search rolls out Gemini 4 Hydrogen globally to all users](https://newruntime.com/posts/google-search-rolls-out-gemini-4-hydrogen-globally-b93679.md)
- 2026-11-14 [Independent check finds Google’s Gemini 4 Hydrogen Legal Research ranking does not match the Vals board](https://newruntime.com/posts/independent-check-finds-google-s-gemini-4-hydrogen-931adc.md)
- 2026-11-14 [Langfuse adds Gemini decision-model evaluators on gemini-decision-1](https://newruntime.com/posts/langfuse-adds-gemini-decision-model-evaluators-on-gemini-2accc5.md)
- 2026-11-14 [Langfuse AI Gateway adds Microsoft Azure Foundry and Google AI Studio](https://newruntime.com/posts/langfuse-ai-gateway-adds-microsoft-azure-foundry-and-64ffdd.md)
- 2026-11-14 [MuJoCo 4.0.1 enables Studio Web Viewer on Windows and fixes a sparse Cholesky bug](https://newruntime.com/posts/mujoco-4-0-1-enables-studio-web-viewer-7afbbf.md)
- 2026-11-14 [Releases: google-deepmind/alphagenome: GitHub](https://newruntime.com/posts/releases-google-deepmind-alphagenome-github-916a1d.md)
- 2026-11-14 [Replit Agent can create Google Ads and buy supplies after user approval](https://newruntime.com/posts/replit-agent-can-create-google-ads-and-buy-aa8f98.md)
- 2026-11-14 [What's new in Gemini 4 Carbon](https://newruntime.com/posts/what-s-new-in-gemini-4-carbon-7ad7fc.md)
- 2026-11-14 [Write-up says HappyDelivery's five reported Cloud Run webhook times all missed the 150 ms goal](https://newruntime.com/posts/write-up-says-happydelivery-s-five-reported-cloud-6f0ea9.md)
- 2026-11-13 [Free My TV 1.2: a ready APK, 26 languages, more TVs](https://newruntime.com/posts/free-my-tv-1-2-a-ready-apk-e03a63.md)
- 2026-11-13 [Intelligent eyewear collections with Gemini features](https://newruntime.com/posts/intelligent-eyewear-collections-with-gemini-features-17b73e.md)
- 2026-11-13 [PARDIG & SatWeave: Foundational Datasets for Earth Science](https://newruntime.com/posts/pardig-satweave-foundational-datasets-for-earth-science-1b4359.md)
- 2026-11-13 [Scientists are using Google AI to help restore giant kelp forests.](https://newruntime.com/posts/scientists-are-using-google-ai-to-help-restore-568f12.md)
- 2026-11-13 [SKU Groups - Google Developer Program Premium GenAI Credit](https://newruntime.com/posts/sku-groups-google-developer-program-premium-genai-credit-529d11.md)
- 2026-11-12 [Google AI Studio added self-hosted sandbox support to Managed Agents on the Gemini API.](https://newruntime.com/posts/google-ai-studio-added-self-hosted-sandbox-support-0322b1.md)
- 2026-11-12 [Google is rolling out Browser Use to all users in Antigravity.](https://newruntime.com/posts/google-is-rolling-out-browser-use-to-all-6d199d.md)
- 2026-11-12 [Introducing Gemini 4 Hydrogen for Agentic Workflows](https://newruntime.com/posts/introducing-gemini-4-hydrogen-for-agentic-workflows-2b9294.md)
- 2026-11-12 [Modal becomes a gVisor maintainer](https://newruntime.com/posts/modal-becomes-a-gvisor-maintainer-05d440.md)
- 2026-11-12 [Pinecone integrated embedding now supports Google's Gemini Embedding 2](https://newruntime.com/posts/pinecone-integrated-embedding-now-supports-google-s-gemini-4e1c0a.md)
- 2026-11-12 [SensorWM is an action-conditioned world model pre-trained on 2.28M person-weeks of Fitbit data.](https://newruntime.com/posts/sensorwm-is-an-action-conditioned-world-model-pre-583b4d.md)
- 2026-11-12 [The Definitive Fine-Tuning Guide for 3D Generation Models](https://newruntime.com/posts/the-definitive-fine-tuning-guide-for-3d-generation-1cd99e.md)
- 2026-11-11 [A2UI 1.0: Build delightful, fast, reliable, agentic experiences](https://newruntime.com/posts/a2ui-1-0-build-delightful-fast-reliable-agentic-456dee.md)
- 2026-11-11 [Alphabet says TPU demand is lifting Cloud while capex and TPU inventory weigh on cash flow](https://newruntime.com/posts/alphabet-says-tpu-demand-is-lifting-cloud-while-91390d.md)
- 2026-11-11 [Chrome releases Modern Web Guidance 1.0 with Firefox and Edge joining the project](https://newruntime.com/posts/chrome-releases-modern-web-guidance-1-0-with-977b2e.md)
- 2026-11-11 [Gemini 4 Carbon is generally available on Google Cloud and in Agent Studio](https://newruntime.com/posts/gemini-4-carbon-is-generally-available-on-google-d0a0ba.md)
- 2026-11-11 [Gemini 4 Carbon ranks #1 on CursorBench 4.0 and Perplexity's WANDR benchmarks for deep agentic work.](https://newruntime.com/posts/gemini-4-carbon-ranks-1-on-cursorbench-4-b4ad54.md)
- 2026-11-11 [Google Antigravity launched /swarm, a preview command orchestrating parallel specialized agents on a shared message board.](https://newruntime.com/posts/google-antigravity-launched-swarm-a-preview-command-orchestrating-434db2.md)
- 2026-11-11 [Google Chrome’s Modern Web Guidance packages expert-curated web-platform recommendations as skills that](https://newruntime.com/posts/google-chrome-s-modern-web-guidance-packages-expert-727856.md)
- 2026-11-11 [Google’s TPU Sync repository describes an open-source C++ library for managing KV caches](https://newruntime.com/posts/google-s-tpu-sync-repository-describes-an-open-169b24.md)
- 2026-11-11 [Much of Broadcom’s increase in shipped compute came from chip volume: Broadcom shipped 65% more chips, or roughly 800,000 in total.](https://newruntime.com/posts/much-of-broadcom-s-increase-in-shipped-compute-532104.md)
- 2026-11-11 [NASA and Google Public Sector test an AI robot for Moon and Mars crew support](https://newruntime.com/posts/nasa-and-google-public-sector-test-an-ai-cd85da.md)
- 2026-11-11 [New AI computing capacity is doubling every 9 months.](https://newruntime.com/posts/new-ai-computing-capacity-is-doubling-every-9-d4c7c1.md)
- 2026-11-11 [On-Device Speech Recognition and Text-to-Speech in Google AI Edge Gallery](https://newruntime.com/posts/on-device-speech-recognition-and-text-to-speech-3f2cdf.md)
- 2026-11-10 [Announcing Long-Context Training on TPUs with MaxText Context Parallelism](https://newruntime.com/posts/announcing-long-context-training-on-tpus-with-maxtext-5a6500.md)
- 2026-11-10 [Antigravity launches a Knowledge Work layout alongside Software Engineering, selectable in Settings > General.](https://newruntime.com/posts/antigravity-launches-a-knowledge-work-layout-alongside-software-03e252.md)
- 2026-11-10 [Gemini 4 Carbon is now available](https://newruntime.com/posts/gemini-4-carbon-is-now-available-4bc61b.md)
- 2026-11-10 [Gemini 4 Carbon is now available for developers.](https://newruntime.com/posts/gemini-4-carbon-is-now-available-for-developers-680123.md)
- 2026-11-10 [Google donates Agent Substrate (1.0) and gVisor to the CNCF as open infrastructure for sandboxed AI agents](https://newruntime.com/posts/google-donates-agent-substrate-1-0-and-gvisor-551fa8.md)
- 2026-11-10 [Meta Muse Ember is listed as available in OpenCode](https://newruntime.com/posts/meta-muse-ember-is-listed-as-available-in-3010fc.md)
- 2026-11-10 [Meta Muse Wonder, Muse Spark 1.3 and Muse Ember now available in Google Cloud model garden](https://newruntime.com/posts/meta-muse-wonder-muse-spark-1-3-and-790011.md)
- 2026-11-10 [MuJoCo 4.0 released with Studio viewer, Filament rendering and deformables](https://newruntime.com/posts/mujoco-4-0-released-with-studio-viewer-filament-cc6d68.md)
- 2026-11-10 [New open-source ASIMOV-Jailbreak benchmark tests adversarial robotics prompts; Argon scores 99.5%.](https://newruntime.com/posts/new-open-source-asimov-jailbreak-benchmark-tests-adversarial-38a3a7.md)
- 2026-11-09 [Developers can now build with Google Maps Platform in AI Studio.](https://newruntime.com/posts/developers-can-now-build-with-google-maps-platform-136d58.md)
- 2026-11-09 [Introducing TPU 8tx and TPU 8i Max](https://newruntime.com/posts/introducing-tpu-8tx-and-tpu-8i-max-b6c8a7.md)
- 2026-11-09 [Sarvam AI names ex Google Cloud India MD Sashikumar Sreedharan as President and CBO, also joining the board.](https://newruntime.com/posts/sarvam-ai-names-ex-google-cloud-india-md-a2ffcd.md)

## GitHub (23)

- 2026-11-15 [VieNeu-TTS v3 Turbo ships 48 kHz Vietnamese and English TTS with instant voice cloning, open-sourced under Apache 2.0 on GitHub](https://newruntime.com/posts/vieneu-tts-v3-turbo-ships-48-khz-vietnamese-138672.md)
- 2026-11-14 [Agentic Workflows v1.2.0 requires opt-in before pull_request_target runs on forks](https://newruntime.com/posts/agentic-workflows-v1-2-0-requires-opt-in-df4b46.md)
- 2026-11-14 [Copilot CLI 1.0.111 adds regex to enterprise-managed shell permission rules](https://newruntime.com/posts/copilot-cli-1-0-111-adds-regex-to-0ce4be.md)
- 2026-11-14 [Copilot CLI 1.0.112 counts resumed-session AI credits toward the session limit](https://newruntime.com/posts/copilot-cli-1-0-112-counts-resumed-session-6b78c3.md)
- 2026-11-14 [GitHub adds SpaceX as a secret scanning partner and tightens false-positive filters](https://newruntime.com/posts/github-adds-spacex-as-a-secret-scanning-partner-b2f0d0.md)
- 2026-11-14 [GitHub Agentic Workflows v1.2.1 adds durable continuation and fail-closed security reviews](https://newruntime.com/posts/github-agentic-workflows-v1-2-1-adds-durable-97a82a.md)
- 2026-11-14 [GitHub Copilot weekly notes add JetBrains Air, Xcode preview, and app merge tools](https://newruntime.com/posts/github-copilot-weekly-notes-add-jetbrains-air-xcode-f47a50.md)
- 2026-11-14 [GitHub public monitoring alerts now have shareable per-alert URLs](https://newruntime.com/posts/github-public-monitoring-alerts-now-have-shareable-per-74ab0a.md)
- 2026-11-14 [GitHub secret scanning now detects Cloudflare Access service tokens](https://newruntime.com/posts/github-secret-scanning-now-detects-cloudflare-access-service-a3ec53.md)
- 2026-11-14 [pull_request_target blocked by default December 2](https://newruntime.com/posts/pull-request-target-blocked-by-default-december-2-fa5b8b.md)
- 2026-11-14 [Visual Studio 2026 18.12 runs Copilot Chat on the Microsoft Copilot SDK by default](https://newruntime.com/posts/visual-studio-2026-18-12-runs-copilot-chat-55cb95.md)
- 2026-11-13 [Vercel Security Review no longer requires a Vercel project to scan your GitHub repos.](https://newruntime.com/posts/vercel-security-review-no-longer-requires-a-vercel-ec5e96.md)
- 2026-11-12 [Ace: GitHub Next's open-source tool lets teams and agents collaborate on the same Mac.](https://newruntime.com/posts/ace-github-next-s-open-source-tool-lets-3659bb.md)
- 2026-11-12 [Build once, run on any agent: A practical guide to agent skills](https://newruntime.com/posts/build-once-run-on-any-agent-a-practical-ac6ca0.md)
- 2026-11-12 [Depi Security chains prompt injection in openai/codex issue labeler into a path to publish the @openai/codex npm package](https://newruntime.com/posts/depi-security-chains-prompt-injection-in-openai-codex-8451f4.md)
- 2026-11-12 [GitHub launches shared Copilot Automations](https://newruntime.com/posts/github-launches-shared-copilot-automations-b9db7d.md)
- 2026-11-11 [Agent Terminal v0.7.0 adds Linux desktop support](https://newruntime.com/posts/agent-terminal-v0-7-0-adds-linux-desktop-698c65.md)
- 2026-11-11 [Claude Code settings.local.json Secrets: 17 months of exposure on GitHub, & Docker Hub](https://newruntime.com/posts/claude-code-settings-local-json-secrets-17-months-9b923f.md)
- 2026-11-11 [Code at github.com/klimzaporojets/emerge, dataset on Hugging Face.](https://newruntime.com/posts/code-at-github-com-klimzaporojets-emerge-dataset-on-206c7c.md)
- 2026-11-11 [codechart: Agent skill + CLI: maps and flowcharts of code changes in terminal, chat and HTML](https://newruntime.com/posts/codechart-agent-skill-cli-maps-and-flowcharts-of-0aa8e7.md)
- 2026-11-10 [Claude Fable 5.2 is generally available in GitHub Copilot](https://newruntime.com/posts/claude-fable-5-2-is-generally-available-in-c85624.md)
- 2026-11-10 [Gemini 4 Carbon in GitHub Copilot](https://newruntime.com/posts/gemini-4-carbon-in-github-copilot-44e8c2.md)
- 2026-11-10 [GitHub Copilot CLI v1.0.109 adds Claude Fable 5.2 to the model picker and CIDR sandbox rules](https://newruntime.com/posts/github-copilot-cli-v1-0-109-adds-claude-c94543.md)

## Anthropic (22)

- 2026-11-14 [AnyLanguageModel 1.3.0 runs a round of tool calls concurrently](https://newruntime.com/posts/anylanguagemodel-1-3-0-runs-a-round-of-aa3324.md)
- 2026-11-14 [FBI FLASH warns of ARTEX, an open-source LLM agent framework weaponized for automated intrusions](https://newruntime.com/posts/fbi-flash-warns-of-artex-an-open-source-3e4b9a.md)
- 2026-11-14 [Issuer Free Writing Prospectus dated November 11, 2026](https://newruntime.com/posts/issuer-free-writing-prospectus-dated-november-11-2026-0f0c10.md)
- 2026-11-13 [Anthropic opens private beta of self-hosted Claude Managed Agents](https://newruntime.com/posts/anthropic-opens-private-beta-of-self-hosted-claude-efc8d5.md)
- 2026-11-13 [Building always-on agents: Claude by Anthropic](https://newruntime.com/posts/building-always-on-agents-claude-by-anthropic-b69b9f.md)
- 2026-11-13 [Building strong safety evaluations \ Anthropic](https://newruntime.com/posts/building-strong-safety-evaluations-anthropic-d8bdac.md)
- 2026-11-13 [OpenAI, Anthropic face new oversight as New York’s AI safety law takes effect](https://newruntime.com/posts/openai-anthropic-face-new-oversight-as-new-york-656343.md)
- 2026-11-12 [Anthropic revealed that Claude Mythos, its most powerful model, moved Claude onto AMD’s chips in three days.](https://newruntime.com/posts/anthropic-revealed-that-claude-mythos-its-most-powerful-438b1c.md)
- 2026-11-12 [Anthropic’s Initial Public Offering](https://newruntime.com/posts/anthropic-s-initial-public-offering-06dea1.md)
- 2026-11-12 [Build composable, interoperable multi-agent systems with A2A on AWS](https://newruntime.com/posts/build-composable-interoperable-multi-agent-systems-with-a2a-379ed1.md)
- 2026-11-12 [Four nonprofits say they have sued Anthropic, Google, Meta, OpenAI and SpaceXSI in San Francisco](https://newruntime.com/posts/four-nonprofits-say-they-have-sued-anthropic-google-2b6db7.md)
- 2026-11-12 [How We Pen Tested LlamaParse Against the OWASP LLM Top 10](https://newruntime.com/posts/how-we-pen-tested-llamaparse-against-the-owasp-455fb9.md)
- 2026-11-12 [Operate Claude Managed Agents inside your security perimeter](https://newruntime.com/posts/operate-claude-managed-agents-inside-your-security-perimeter-119ad1.md)
- 2026-11-11 [Anthropic discloses $10 billion commitment to Google, Microsoft](https://newruntime.com/posts/anthropic-discloses-10-billion-commitment-to-google-microsoft-21f156.md)
- 2026-11-11 [Anthropic has started the roadshow for a proposed initial public offering of 825 million shares at an](https://newruntime.com/posts/anthropic-has-started-the-roadshow-for-a-proposed-9c8171.md)
- 2026-11-11 [Goldman Sachs expects Anthropic revenue to hit $600 billion by 2030](https://newruntime.com/posts/goldman-sachs-expects-anthropic-revenue-to-hit-600-50f47d.md)
- 2026-11-11 [One contributor to recent global compute growth is Broadcom, which produces Google’s custom AI chips (TPUs).](https://newruntime.com/posts/one-contributor-to-recent-global-compute-growth-is-15b155.md)
- 2026-11-10 [Anthropic launches roadshow for proposed initial public offering \ Anthropic](https://newruntime.com/posts/anthropic-launches-roadshow-for-proposed-initial-public-offering-f890ca.md)
- 2026-11-09 [Anthropic IPO valuation faces skepticism as investors question $1.5 trillion](https://newruntime.com/posts/anthropic-ipo-valuation-faces-skepticism-as-investors-question-218cce.md)
- 2026-11-09 [Anthropic strikes $3.4bn deal to generate gas-fired power for data centres](https://newruntime.com/posts/anthropic-strikes-3-4bn-deal-to-generate-gas-0d7e4c.md)
- 2026-11-09 [Ramp’s September 2026 index says 55.9% of businesses in its U.S](https://newruntime.com/posts/ramp-s-september-2026-index-says-55-9-73ac57.md)
- 2026-11-09 [Scientists curb AI use after companies announced discoveries researchers had pursued with those tools](https://newruntime.com/posts/scientists-curb-ai-use-after-companies-announced-discoveries-15514b.md)

## Inference (22)

- 2026-11-15 [Release v0.7.0: huggingface/lerobot](https://newruntime.com/posts/release-v0-7-0-huggingface-lerobot-2919c9.md)
- 2026-11-14 [NVIDIA adds official SGLang instructions for the NVFP4 DeepSeek checkpoint](https://newruntime.com/posts/nvidia-adds-official-sglang-instructions-for-the-nvfp4-00fbd0.md)
- 2026-11-13 [DeepSeek-V4.1-Flash on Vera Rubin NVL72: 58x Throughput vs. GB300 NVL72 at 387 tokens/s](https://newruntime.com/posts/deepseek-v4-1-flash-on-vera-rubin-nvl72-0d61a3.md)
- 2026-11-13 [disaggregated-quantization: Disaggregated quantization official codebase](https://newruntime.com/posts/disaggregated-quantization-disaggregated-quantization-official-codebase-3c1bec.md)
- 2026-11-13 [Eliminating Nondeterminism in LLM Inference with Integer-Only Arithmetic](https://newruntime.com/posts/eliminating-nondeterminism-in-llm-inference-with-integer-only-1ff8b2.md)
- 2026-11-13 [Lightning Labs ships pi402, a Pi coding agent extension that pays for LLM inference over L402](https://newruntime.com/posts/lightning-labs-ships-pi402-a-pi-coding-agent-c76794.md)
- 2026-11-13 [Ships Q4_0, Q4_K_M, Q8_0 GGUF builds for llama.cpp on-device inference.](https://newruntime.com/posts/ships-q4-0-q4-k-m-q8-0-c575d2.md)
- 2026-11-13 [Two new hyperscaler wins including a top-three cloud, with Oracle integration in 2027](https://newruntime.com/posts/two-new-hyperscaler-wins-including-a-top-three-683d2e.md)
- 2026-11-12 [Day-0 Support for Flint on vLLM](https://newruntime.com/posts/day-0-support-for-flint-on-vllm-2c6d85.md)
- 2026-11-12 [hs-rerank-1 is a multilingual cross-encoder reranker with 114M params, MIT licensed.](https://newruntime.com/posts/hs-rerank-1-is-a-multilingual-cross-encoder-bb44a6.md)
- 2026-11-12 [oMLX 0.8.0rc1 adds Prefill Accelerator using NVIDIA GPUs or a second Mac](https://newruntime.com/posts/omlx-0-8-0rc1-adds-prefill-accelerator-using-780453.md)
- 2026-11-11 [Muse Spark Open Ultrafast: Pushing beyond 1,000 Tokens per second](https://newruntime.com/posts/muse-spark-open-ultrafast-pushing-beyond-1-000-fe17eb.md)
- 2026-11-11 [Runway releases Praxis-1, a 27B open-weight world action model for robots, under Apache 2.0](https://newruntime.com/posts/runway-releases-praxis-1-a-27b-open-weight-36734b.md)
- 2026-11-11 [Ships with inference code, a 5M-sample dataset, and ComfyUI plus GGUF community ports.](https://newruntime.com/posts/ships-with-inference-code-a-5m-sample-dataset-5c0691.md)
- 2026-11-11 [The platform for inference, training, and execution](https://newruntime.com/posts/the-platform-for-inference-training-and-execution-896444.md)
- 2026-11-10 [Epoch AI launched the Inference Performance Index, scoring chips by real inference capacity, not FLOP/s.](https://newruntime.com/posts/epoch-ai-launched-the-inference-performance-index-scoring-1bb9b1.md)
- 2026-11-10 [NEAR AI incident report: SGLang remote code execution let attacker into inference servers, including requests processed in confidential mode](https://newruntime.com/posts/near-ai-incident-report-sglang-remote-code-execution-46e998.md)
- 2026-11-10 [NVIDIA releases open-source Snapshot for checkpoint and restore of GPU inference workloads on Kubernetes](https://newruntime.com/posts/nvidia-releases-open-source-snapshot-for-checkpoint-and-57b093.md)
- 2026-11-10 [Update on October 2026 security incident](https://newruntime.com/posts/update-on-october-2026-security-incident-c8af72.md)
- 2026-11-09 [Baidu ERNIE Evolve: automated research system beats public NanoChat and NanoGPT Speedrun records and speeds up ERNIE-5.5-Flash inference](https://newruntime.com/posts/baidu-ernie-evolve-automated-research-system-beats-public-70dab4.md)
- 2026-11-09 [Control Room: A central system to track and stop model workloads across our research infrastructure](https://newruntime.com/posts/control-room-a-central-system-to-track-and-38b953.md)
- 2026-11-09 [InferenceX measures Muse Spark Open on GB300 NVL72 at 43,643 tok/s per GPU and $0.015/M tokens](https://newruntime.com/posts/inferencex-measures-muse-spark-open-on-gb300-nvl72-e05cc5.md)

## Agents (20)

- 2026-11-14 [LangChain Academy New Course: Building and Deploying Managed Agents](https://newruntime.com/posts/langchain-academy-new-course-building-and-deploying-managed-bfa1d7.md)
- 2026-11-14 [LangChain’s account of Actively AI describes a redesign of sales agents](https://newruntime.com/posts/langchain-s-account-of-actively-ai-describes-a-3c9f73.md)
- 2026-11-14 [Meta lists Muse from Meta Chrome extension for the browser side panel](https://newruntime.com/posts/meta-lists-muse-from-meta-chrome-extension-for-6024bd.md)
- 2026-11-14 [Sesame voice agent is ending its "preview" period and introducing Plus and Pro paid plans.](https://newruntime.com/posts/sesame-voice-agent-is-ending-its-preview-period-20d8ce.md)
- 2026-11-13 [Adding Teammates to the native client places agent controls alongside Perplexity’s voice mode and local Mac mode.](https://newruntime.com/posts/adding-teammates-to-the-native-client-places-agent-500be4.md)
- 2026-11-13 [Build a Robot Pick-and-Place Simulation with SimReady Assets](https://newruntime.com/posts/build-a-robot-pick-and-place-simulation-with-90f9e1.md)
- 2026-11-13 [Cognition launched Devin Apps, private web applications hosted on Devin's cloud VM with stable URLs.](https://newruntime.com/posts/cognition-launched-devin-apps-private-web-applications-hosted-65bf01.md)
- 2026-11-13 [How Natera designed an agentic AI system for lab ops with Amazon Bedrock AgentCore](https://newruntime.com/posts/how-natera-designed-an-agentic-ai-system-for-378cd2.md)
- 2026-11-13 [Krea opened the beta for Krea Bots, named AI agents that each own one creative craft.](https://newruntime.com/posts/krea-opened-the-beta-for-krea-bots-named-39050d.md)
- 2026-11-13 [LangChain Raises $300M Series C at $3B Valuation to Advance AI Agents](https://newruntime.com/posts/langchain-raises-300m-series-c-at-3b-valuation-e41e99.md)
- 2026-11-13 [OpenAI is rolling out ChatGPT Wallet in dots](https://newruntime.com/posts/openai-is-rolling-out-chatgpt-wallet-in-dots-1e24aa.md)
- 2026-11-12 [The Latest in Agent Traffic: Meta Muse Starts Signing Its Traffic and Grok Bot Changes Its Route](https://newruntime.com/posts/the-latest-in-agent-traffic-meta-muse-starts-45dc4b.md)
- 2026-11-12 [United won’t let your AI agent book your flight](https://newruntime.com/posts/united-won-t-let-your-ai-agent-book-e21f2f.md)
- 2026-11-11 [How our GTM Agent builds consistent, high-quality sales artifacts](https://newruntime.com/posts/how-our-gtm-agent-builds-consistent-high-quality-ce46b3.md)
- 2026-11-11 [n8n Labs releases Nod, an open-source personal agent that runs on the user's computer](https://newruntime.com/posts/n8n-labs-releases-nod-an-open-source-personal-31bb52.md)
- 2026-11-10 [Agent now writes Python in a persistent REPL running directly on Chrome DevTools Protocol.](https://newruntime.com/posts/agent-now-writes-python-in-a-persistent-repl-f84703.md)
- 2026-11-09 [DoubleAI says it placed an AI agent inside neural-network training runs and let it intervene in what the](https://newruntime.com/posts/doubleai-says-it-placed-an-ai-agent-inside-c0ce7d.md)
- 2026-11-09 [Introducing ElevenAgents Campaign Manager](https://newruntime.com/posts/introducing-elevenagents-campaign-manager-083d89.md)
- 2026-11-09 [OpenClaw v2026.11.2 patch is out!](https://newruntime.com/posts/openclaw-v2026-11-2-patch-is-out-54e8e7.md)
- 2026-11-09 [Teammates are agents for complex, ongoing work](https://newruntime.com/posts/teammates-are-agents-for-complex-ongoing-work-8c62d9.md)

## MCP (18)

- 2026-11-14 [67.js is an MIT-licensed WebGL2 3D engine built for AI agents to drive.](https://newruntime.com/posts/67-js-is-an-mit-licensed-webgl2-3d-cf8319.md)
- 2026-11-14 [Factory’s November 14, 2026 changelog lists CLI and Desktop v1.12.0, with changes to tool-call visibility](https://newruntime.com/posts/factory-s-november-14-2026-changelog-lists-cli-2fc46a.md)
- 2026-11-14 [Frigade released Yogurt, an MIT-licensed local AI meeting notes app.](https://newruntime.com/posts/frigade-released-yogurt-an-mit-licensed-local-ai-578386.md)
- 2026-11-13 [opencode v2.3.0 makes subagents always run in the background and adds a quieter TUI transcript](https://newruntime.com/posts/opencode-v2-3-0-makes-subagents-always-run-0c81a8.md)
- 2026-11-13 [Release v1.28.0-pre: zed-industries/zed](https://newruntime.com/posts/release-v1-28-0-pre-zed-industries-zed-8ef2ff.md)
- 2026-11-12 [Genie Ontology, now generally available, supplies the structured business context behind the agent.](https://newruntime.com/posts/genie-ontology-now-generally-available-supplies-the-structured-e97e90.md)
- 2026-11-12 [MCP vs CLI vs Skills: What Each Costs in Context](https://newruntime.com/posts/mcp-vs-cli-vs-skills-what-each-costs-f3de5f.md)
- 2026-11-12 [OpenAI open-sources MCP Extensions for ChatGPT-native plugins](https://newruntime.com/posts/openai-open-sources-mcp-extensions-for-chatgpt-native-672594.md)
- 2026-11-11 [A More Predictable Release Cycle for MCP](https://newruntime.com/posts/a-more-predictable-release-cycle-for-mcp-d720a5.md)
- 2026-11-11 [earshot v0.4.1: community fixes](https://newruntime.com/posts/earshot-v0-4-1-community-fixes-f565f3.md)
- 2026-11-11 [Epic ships Unreal Engine 5.9 MegaJam Preview, built mainly with agentic workflows](https://newruntime.com/posts/epic-ships-unreal-engine-5-9-megajam-preview-62f989.md)
- 2026-11-11 [Exploring File Uploads and Downloads in MCP](https://newruntime.com/posts/exploring-file-uploads-and-downloads-in-mcp-98cc90.md)
- 2026-11-11 [New in Delta: your thread can keep running while you touch grass.](https://newruntime.com/posts/new-in-delta-your-thread-can-keep-running-1a15ff.md)
- 2026-11-11 [Parallel Search is now available in AWS Marketplace](https://newruntime.com/posts/parallel-search-is-now-available-in-aws-marketplace-821d82.md)
- 2026-11-11 [Public exploit published for unauthenticated RCE in Ruflo MCP Bridge 3.16.2](https://newruntime.com/posts/public-exploit-published-for-unauthenticated-rce-in-ruflo-c3ef96.md)
- 2026-11-09 [Red Hat OpenShift 5 reaches GA with a generally available MCP server for AI agents](https://newruntime.com/posts/red-hat-openshift-5-reaches-ga-with-a-ef447a.md)
- 2026-11-09 [Vercel MCP is now open to any MCP client.](https://newruntime.com/posts/vercel-mcp-is-now-open-to-any-mcp-5ceef2.md)
- 2026-11-09 [Vercel MCP now publishes events.](https://newruntime.com/posts/vercel-mcp-now-publishes-events-ecc542.md)

## Gemini (17)

- 2026-11-14 [AnyLanguageModel 1.2.0 adds dynamic profiles and changes tool loops](https://newruntime.com/posts/anylanguagemodel-1-2-0-adds-dynamic-profiles-and-3da9ab.md)
- 2026-11-14 [Gemini Omni 1.2 Flash: Gemini Enterprise Agent Platform](https://newruntime.com/posts/gemini-omni-1-2-flash-gemini-enterprise-agent-e8e959.md)
- 2026-11-14 [Using a Smart Clip Router to Benchmark Conversational Speech-To-Speech Agents Engineering LLM](https://newruntime.com/posts/using-a-smart-clip-router-to-benchmark-conversational-b36d84.md)
- 2026-11-13 [AI chatbots are recommending fake retail websites, study finds](https://newruntime.com/posts/ai-chatbots-are-recommending-fake-retail-websites-study-f46b3e.md)
- 2026-11-13 [Catch Go bugs that fuzzing misses with Gemini and a constraint solver](https://newruntime.com/posts/catch-go-bugs-that-fuzzing-misses-with-gemini-d9550c.md)
- 2026-11-13 [Model versions and lifecycle: Gemini Enterprise Agent Platform](https://newruntime.com/posts/model-versions-and-lifecycle-gemini-enterprise-agent-platform-6962a5.md)
- 2026-11-13 [Release 0.36: simonw/llm-gemini](https://newruntime.com/posts/release-0-36-simonw-llm-gemini-7adbfa.md)
- 2026-11-13 [Simon Willison’s llm-gemini 0.37 adds support for Google’s Gemini 4 Hydrogen](https://newruntime.com/posts/simon-willison-s-llm-gemini-0-37-adds-b83a88.md)
- 2026-11-12 [Gemini 4 Hydrogen is now available in Cursor!](https://newruntime.com/posts/gemini-4-hydrogen-is-now-available-in-cursor-0da1b8.md)
- 2026-11-12 [Gemini 4 Hydrogen now available on AI Gateway](https://newruntime.com/posts/gemini-4-hydrogen-now-available-on-ai-gateway-7b3652.md)
- 2026-11-11 [Gemini 4 Carbon now live in Perplexity Computer for Pro and Max, and in Search for Max.](https://newruntime.com/posts/gemini-4-carbon-now-live-in-perplexity-computer-18273d.md)
- 2026-11-11 [Gemini Decisions API: Fast multimodal decisions](https://newruntime.com/posts/gemini-decisions-api-fast-multimodal-decisions-b7faa0.md)
- 2026-11-11 [Introducing Gemini for healthcare](https://newruntime.com/posts/introducing-gemini-for-healthcare-d4acd8.md)
- 2026-11-11 [Models - Gemini API](https://newruntime.com/posts/models-gemini-api-1ba0b4.md)
- 2026-11-10 [Gemini 4 Carbon is live in Devin Desktop and CLI, scoring 40.0% on FrontierCode 2.0.](https://newruntime.com/posts/gemini-4-carbon-is-live-in-devin-desktop-f5900a.md)
- 2026-11-10 [Gemini Robotics 2.5: Accelerating Physical Intelligence](https://newruntime.com/posts/gemini-robotics-2-5-accelerating-physical-intelligence-d42905.md)
- 2026-11-10 [Reflection Mirror CLI 0.1.1 adds /fork, /web-provider, and sub-agent inspection](https://newruntime.com/posts/reflection-mirror-cli-0-1-1-adds-fork-2926c6.md)

## Security (17)

- 2026-11-14 [ARTEX and beyond: AI-enabled penetration testing platforms](https://newruntime.com/posts/artex-and-beyond-ai-enabled-penetration-testing-platforms-73d67a.md)
- 2026-11-14 [Node.js 26.13.1 (LTS) "Lithium"](https://newruntime.com/posts/node-js-26-13-1-lts-lithium-fd17ff.md)
- 2026-11-14 [OpenAI is introducing ChatGPT Wallet](https://newruntime.com/posts/openai-is-introducing-chatgpt-wallet-7ec879.md)
- 2026-11-14 [OpenAI's ChatGPT Wallet allows 'dots' AI agents to make purchases](https://newruntime.com/posts/openai-s-chatgpt-wallet-allows-dots-ai-agents-91ef96.md)
- 2026-11-14 [OpenShell v0.1.7 tags the fixed OIDC callback port for container CLI login](https://newruntime.com/posts/openshell-v0-1-7-tags-the-fixed-oidc-13d24f.md)
- 2026-11-14 [Semafor: Nvidia's Huang used his Trump relationship to shape China chip-export rules](https://newruntime.com/posts/semafor-nvidia-s-huang-used-his-trump-relationship-82bf5b.md)
- 2026-11-12 [Cyber Chaos Monkey: Staying ahead of security threats with AI agents](https://newruntime.com/posts/cyber-chaos-monkey-staying-ahead-of-security-threats-f28704.md)
- 2026-11-12 [Introducing the Cyber Defense Network: an ecosystem of cybersecurity partners connected with our Security Factory platform.](https://newruntime.com/posts/introducing-the-cyber-defense-network-an-ecosystem-of-22e6e7.md)
- 2026-11-12 [Turning your social media feed into bait.](https://newruntime.com/posts/turning-your-social-media-feed-into-bait-a86166.md)
- 2026-11-11 [Claroc says 35,578 Grok Build apps opened without sign-in, with 650 to 800 holding private data (TeslaNorth)](https://newruntime.com/posts/claroc-says-35-578-grok-build-apps-opened-990a80.md)
- 2026-11-11 [Security of Cross-Device Flows](https://newruntime.com/posts/security-of-cross-device-flows-1f93c7.md)
- 2026-11-10 [Binance confirms security incident, pauses withdrawals, estimates $6.3B impact](https://newruntime.com/posts/binance-confirms-security-incident-pauses-withdrawals-estimates-6-d8eb12.md)
- 2026-11-10 [Factory has introduced Security Factory, a private-preview product that links security signals](https://newruntime.com/posts/factory-has-introduced-security-factory-a-private-preview-3b1977.md)
- 2026-11-10 [Polymarket says a compromised third-party script drained some user wallets](https://newruntime.com/posts/polymarket-says-a-compromised-third-party-script-drained-5fbb8f.md)
- 2026-11-09 [GPT-6 Cyber System Card](https://newruntime.com/posts/gpt-6-cyber-system-card-045495.md)
- 2026-11-09 [Introducing Control Room: Human control over AI workloads](https://newruntime.com/posts/introducing-control-room-human-control-over-ai-workloads-7841e4.md)
- 2026-11-09 [Vercel Shop 2026.10.1 moves storefronts to Next.js 16.4.3 to patch two critical RCEs](https://newruntime.com/posts/vercel-shop-2026-10-1-moves-storefronts-to-03d1ea.md)

## Codex (14)

- 2026-11-15 [agentes.chat is an AGPL-3.0 self-hosted Slack alternative where AI agents are workspace members.](https://newruntime.com/posts/agentes-chat-is-an-agpl-3-0-self-fc2c7d.md)
- 2026-11-14 [avenoxbeyin gives Claude Code, Codex, and Antigravity persistent cross-session memory via an Obsidian vault](https://newruntime.com/posts/avenoxbeyin-gives-claude-code-codex-and-antigravity-persistent-5a0e74.md)
- 2026-11-14 [Codex CLI 0.169.0 adds side-chat tabs, command-center search, and custom providers in the terminal](https://newruntime.com/posts/codex-cli-0-169-0-adds-side-chat-58c02a.md)
- 2026-11-13 [Codex CLI 0.168.0 changelog lists Windows MXC sandbox path and glob changes](https://newruntime.com/posts/codex-cli-0-168-0-changelog-lists-windows-6f84db.md)
- 2026-11-13 [Securing persistent agents like OpenClaw at OpenAI](https://newruntime.com/posts/securing-persistent-agents-like-openclaw-at-openai-0fdd67.md)
- 2026-11-12 [Create your dot in the ChatGPT mobile app](https://newruntime.com/posts/create-your-dot-in-the-chatgpt-mobile-app-24bbf3.md)
- 2026-11-12 [Reason Network v0.12.13 release](https://newruntime.com/posts/reason-network-v0-12-13-release-f9b958.md)
- 2026-11-11 [AIPOCH Open-Science v1.2.1 adds Codex subscriptions and Claude Fable 5.2 to its research agent](https://newruntime.com/posts/aipoch-open-science-v1-2-1-adds-codex-d575c8.md)
- 2026-11-11 [Oracle expands OpenAI partnership with AI Workspace inside ChatGPT](https://newruntime.com/posts/oracle-expands-openai-partnership-with-ai-workspace-inside-386fe3.md)
- 2026-11-10 [Codex spends thousands of tokens on the default skill catalog before the first user message](https://newruntime.com/posts/codex-spends-thousands-of-tokens-on-the-default-fa4d7a.md)
- 2026-11-10 [Dots on ChatGPT can now connect to multiple local computers, so users can tell them where to run the task.](https://newruntime.com/posts/dots-on-chatgpt-can-now-connect-to-multiple-ffaa7b.md)
- 2026-11-10 [ModRetro Chromatic: Codex Edition](https://newruntime.com/posts/modretro-chromatic-codex-edition-openai-supply-co-620729.md)
- 2026-11-10 [OpenAI doubles ChatGPT Pro usage limits after backlash over planned 50% cut](https://newruntime.com/posts/openai-doubles-chatgpt-pro-usage-limits-after-backlash-965f5e.md)
- 2026-11-09 [OpenAI is teasing ModRetro Chromatic to arrive at Supply Co.](https://newruntime.com/posts/openai-is-teasing-modretro-chromatic-to-arrive-at-20f239.md)

## AI infrastructure (13)

- 2026-11-14 [Notion says Fireworks attackers took Insights chat transcripts from under 1% of workspaces](https://newruntime.com/posts/notion-says-fireworks-attackers-took-insights-chat-transcripts-a616da.md)
- 2026-11-14 [Nscale files S-1/A estimating $307.5 million nine-month revenue and $1.69 billion net loss](https://newruntime.com/posts/nscale-files-s-1-a-estimating-307-5-50a5c4.md)
- 2026-11-12 [Georgia residents file civil RICO suit against OpenAI and Effingham County officials over Project Camellia](https://newruntime.com/posts/georgia-residents-file-civil-rico-suit-against-openai-6b9107.md)
- 2026-11-12 [Huawei, China’s biggest AI chip developer, is drawing on China’s abundant energy to build data centers.](https://newruntime.com/posts/huawei-china-s-biggest-ai-chip-developer-is-82f78f.md)
- 2026-11-12 [Meta’s New Agreements Bring Clean Manufacturing to Data Center Supply Chains](https://newruntime.com/posts/meta-s-new-agreements-bring-clean-manufacturing-to-0e436c.md)
- 2026-11-11 [CoreWeave Q3 2026: revenue up 168% to $3.66B, backlog $132.5B, full-year guidance raised](https://newruntime.com/posts/coreweave-q3-2026-revenue-up-168-to-3-4521ab.md)
- 2026-11-11 [Elon Musk warns 200-plus neoclouds of guppy-level death as shares fall](https://newruntime.com/posts/elon-musk-warns-200-plus-neoclouds-of-guppy-1882c6.md)
- 2026-11-11 [Inside The Weed Industry’s AI Data Center Pipe Dream](https://newruntime.com/posts/inside-the-weed-industry-s-ai-data-center-06a540.md)
- 2026-11-09 [Microsoft Sours on Neoclouds IREN, Nebius, Nscale on Delays, Costs](https://newruntime.com/posts/microsoft-sours-on-neoclouds-iren-nebius-nscale-on-84e031.md)
- 2026-11-09 [NVIDIA Announces New Planned Investments in Korean Startups at the MSIT-NVIDIA AI Startup Partnership Forum](https://newruntime.com/posts/nvidia-announces-new-planned-investments-in-korean-startups-1f8257.md)
- 2026-11-09 [OpenAI will keep investing in AI data centers despite election results, CFO tells CNBC](https://newruntime.com/posts/openai-will-keep-investing-in-ai-data-centers-0ab46c.md)
- 2026-11-09 [The Election Day Revolt That Threatens America’s AI Boom](https://newruntime.com/posts/the-election-day-revolt-that-threatens-america-s-8a7524.md)
- 2026-11-09 [TorchTitan-Zephon bumped to Zephon 0.2 with TCPStore checkpoint aggregation](https://newruntime.com/posts/torchtitan-zephon-bumped-to-zephon-0-2-with-4961a4.md)

## Coding agents (11)

- 2026-11-14 [Artificial Analysis publishes Coding Agent Index v2.1 with FrontierCode-AA 2.0, Terminal-Bench 4.1 and SWE-Atlas-QnA-AA](https://newruntime.com/posts/artificial-analysis-publishes-coding-agent-index-v2-1-983801.md)
- 2026-11-14 [Caveman is an open-source CLI that changes how coding agents handle responses, file output](https://newruntime.com/posts/caveman-is-an-open-source-cli-that-changes-23b835.md)
- 2026-11-13 [Gumloop says its internal Software Factory merged 571 agent-written pull requests in six weeks](https://newruntime.com/posts/gumloop-says-its-internal-software-factory-merged-571-68e712.md)
- 2026-11-12 [Microsoft publishes two studies on post-training supply chain backdoors in coding agents](https://newruntime.com/posts/microsoft-publishes-two-studies-on-post-training-supply-edd99e.md)
- 2026-11-11 [Orchestrating a One-Person Software Factory: run many coding agents without becoming their PM](https://newruntime.com/posts/orchestrating-a-one-person-software-factory-run-many-5705c0.md)
- 2026-11-10 [Berkeley 'Evolving Agent Teams': supervisor-restructured coding agents beat FlashInfer MLA decode kernel by 1.21x](https://newruntime.com/posts/berkeley-evolving-agent-teams-supervisor-restructured-coding-agents-4a7472.md)
- 2026-11-10 [How we rebuilt the Shopify app in 12 weeks](https://newruntime.com/posts/how-we-rebuilt-the-shopify-app-in-12-c85550.md)
- 2026-11-10 [Particle-1 is a coding model trained to finish tasks with fewer tokens](https://newruntime.com/posts/particle-1-is-a-coding-model-trained-to-fe301e.md)
- 2026-11-09 [print3d: The best skill for 3D printing](https://newruntime.com/posts/print3d-the-best-skill-for-3d-printing-071dd7.md)
- 2026-11-09 [Pylon x Devin: From bug report to code fix](https://newruntime.com/posts/pylon-x-devin-from-bug-report-to-code-a5809f.md)
- 2026-11-09 [vork is a headless Rust/Slang 3D engine where coding agents write scenes as code and render on GPU.](https://newruntime.com/posts/vork-is-a-headless-rust-slang-3d-engine-12ecec.md)

## Generative media (11)

- 2026-11-14 [ACE-Step 1.5 is an open-source text-to-music foundation model from ACE Studio and StepFun, available on Hugging Face under Apache-2.0.](https://newruntime.com/posts/ace-step-1-5-is-an-open-source-5736d6.md)
- 2026-11-14 [Natively integrated in ComfyUI via Template Library → Audio with an all-in-one workflow.](https://newruntime.com/posts/natively-integrated-in-comfyui-via-template-library-audio-6d4b45.md)
- 2026-11-14 [Physicl raises $10M to make the digital world real enough for robots](https://newruntime.com/posts/physicl-raises-10m-to-make-the-digital-world-5aadb9.md)
- 2026-11-14 [Suno launches Bridge, a standalone Mac app that generates audio into a DAW](https://newruntime.com/posts/suno-launches-bridge-a-standalone-mac-app-that-64673d.md)
- 2026-11-13 [Scammers using AI-generated images of missing pets to demand reward money](https://newruntime.com/posts/scammers-using-ai-generated-images-of-missing-pets-a49cf2.md)
- 2026-11-12 [earshot 0.4.0 adds 51 recipe-made sounds, a launch kit preset, and a remade fit engine](https://newruntime.com/posts/earshot-0-4-0-adds-51-recipe-made-d183bf.md)
- 2026-11-11 [Black Forest Labs docs added a page for flux-3-max](https://newruntime.com/posts/black-forest-labs-docs-added-a-page-for-5a11f6.md)
- 2026-11-11 [Comfy-Org published a repackaged, quantized MiniMax H3-Image build ready to drop into ComfyUI](https://newruntime.com/posts/comfy-org-published-a-repackaged-quantized-minimax-h3-6fa189.md)
- 2026-11-09 [Answer to Complaint – #11 in UMG Recordings, Inc. v. Suno, Inc. (D. Mass., 1:26-cv-14275)](https://newruntime.com/posts/answer-to-complaint-11-in-umg-recordings-inc-9557dc.md)
- 2026-11-09 [Research prototype streams a live 3D Gaussian Splatting scene instead of rendered video frames.](https://newruntime.com/posts/research-prototype-streams-a-live-3d-gaussian-splatting-f7477f.md)
- 2026-11-09 [Seed News - ByteDance Seed Team](https://newruntime.com/posts/seed-news-bytedance-seed-team-bc7481.md)

## Compute (8)

- 2026-11-13 [Epoch AI releases AI chip imports explorer](https://newruntime.com/posts/epoch-ai-releases-ai-chip-imports-explorer-4bb8c3.md)
- 2026-11-12 [At 81,000 H100-equivalents, Huawei Horinger is a large AI data center by today’s standards, and probably one of the largest in China.](https://newruntime.com/posts/at-81-000-h100-equivalents-huawei-horinger-is-b6f1d8.md)
- 2026-11-12 [Don't Write Kernels, Derive Them](https://newruntime.com/posts/don-t-write-kernels-derive-them-aee769.md)
- 2026-11-11 [AMD launches 10-CPU Epyc 6 'Venice' line for AI agent workloads, claims 3.3x more agents per watt than Nvidia Vera](https://newruntime.com/posts/amd-launches-10-cpu-epyc-6-venice-line-8035c3.md)
- 2026-11-11 [What Huawei’s Horinger data center reveals about China’s compute gap](https://newruntime.com/posts/what-huawei-s-horinger-data-center-reveals-about-f694cb.md)
- 2026-11-11 [Where compute ends up, country by country.](https://newruntime.com/posts/where-compute-ends-up-country-by-country-71a4d2.md)
- 2026-11-10 [Altman says OpenAI will point robotics at data centers; self-replicating sites are within grasp](https://newruntime.com/posts/altman-says-openai-will-point-robotics-at-data-54dcf5.md)
- 2026-11-10 [NVIDIA Vera brings faster compute to E2B sandboxes](https://newruntime.com/posts/nvidia-vera-brings-faster-compute-to-e2b-sandboxes-795676.md)

## Multimodal (7)

- 2026-11-14 [Artificial Analysis launched a Diarization Index ranking speech to text APIs on Word Diarization Error Rate](https://newruntime.com/posts/artificial-analysis-launched-a-diarization-index-ranking-speech-f7f931.md)
- 2026-11-14 [Inkling 2 Small Model Card](https://newruntime.com/posts/inkling-2-small-model-card-826206.md)
- 2026-11-14 [Production selection also depends on turnaround time and cost.](https://newruntime.com/posts/production-selection-also-depends-on-turnaround-time-and-546a1f.md)
- 2026-11-12 [NVIDIA released Déjà View, a 117M-parameter multi-view 3D reconstruction transformer, NeurIPS 2026 oral.](https://newruntime.com/posts/nvidia-released-deja-view-a-117m-parameter-multi-8bc008.md)
- 2026-11-11 [Indra OCR 1.1.1 and 1.1.2 fix visual-order RTL PDFs and legacy-font Hindi text layers](https://newruntime.com/posts/indra-ocr-1-1-1-and-1-1-63fe4c.md)
- 2026-11-11 [Release Indra OCR 1.1.0](https://newruntime.com/posts/release-indra-ocr-1-1-0-datalab-to-ac3594.md)
- 2026-11-10 [ByteDance released Seedream 5.5, now live on Higgsfield ahead of US rollout.](https://newruntime.com/posts/bytedance-released-seedream-5-5-now-live-on-89c03e.md)

## Open source (7)

- 2026-11-13 [Kitesurf is now open source](https://newruntime.com/posts/kitesurf-is-now-open-source-19d104.md)
- 2026-11-12 [Drift Inspector: Exploring and Measuring Scientific Drift with Atomic Contribution Claims](https://newruntime.com/posts/drift-inspector-exploring-and-measuring-scientific-drift-with-582540.md)
- 2026-11-11 [Skipper, Airbnb’s embedded workflow engine, is now open source: Airbnb Engineering & Data Science](https://newruntime.com/posts/skipper-airbnb-s-embedded-workflow-engine-is-now-9625b1.md)
- 2026-11-10 [NXP launches unified robotics platform with open-source SDK and drone reference designs](https://newruntime.com/posts/nxp-launches-unified-robotics-platform-with-open-source-274a25.md)
- 2026-11-09 [Cloudflare is distributing the first $301,300 of the Vite Ecosystem Fund to open source projects and maintainers.](https://newruntime.com/posts/cloudflare-is-distributing-the-first-301-300-of-e8e70a.md)
- 2026-11-09 [Nemotron-Personas-Finance-Japan: A Synthetic Dataset Capturing Japan's Demographics and Financial Context](https://newruntime.com/posts/nemotron-personas-finance-japan-a-synthetic-dataset-capturing-f53831.md)
- 2026-11-09 [Skild AI and NVIDIA open-source Steel, a GPU rigid-body solver for Newton](https://newruntime.com/posts/skild-ai-and-nvidia-open-source-steel-a-046417.md)

## Agent evals (6)

- 2026-11-12 [ParseBench-2: Severity-aware document parsing benchmark](https://newruntime.com/posts/parsebench-2-severity-aware-document-parsing-benchmark-fa4acb.md)
- 2026-11-11 [ACE Music v1 debuts at #1 on AA-Music-Vocal v1.1 with Elo 1160, ahead of Suno v6 at 1144.](https://newruntime.com/posts/ace-music-v1-debuts-at-1-on-aa-8eac84.md)
- 2026-11-11 [Scale Labs Leaderboard: Remote Labor Index (RLI)](https://newruntime.com/posts/scale-labs-leaderboard-remote-labor-index-rli-4e7b47.md)
- 2026-11-11 [UAEF: A framework-agnostic system for closed-loop evaluation of AI agents](https://newruntime.com/posts/uaef-a-framework-agnostic-system-for-closed-loop-c647d5.md)
- 2026-11-10 [Reported 68.0% on FinSearchComp, though third-party tests show competitors leading on T3 complex tasks](https://newruntime.com/posts/reported-68-0-on-finsearchcomp-though-third-party-07c21e.md)
- 2026-11-09 [What Is Eval-Driven Development? How to Improve AI Agents](https://newruntime.com/posts/what-is-eval-driven-development-how-to-improve-52927d.md)

## Agent harness (6)

- 2026-11-14 [DeepSeek Harness dsh-v0.2.2-rc.3 adds experimental DeepSeek chat import](https://newruntime.com/posts/deepseek-harness-dsh-v0-2-2-rc-3-72b70c.md)
- 2026-11-14 [Hugging Face wires safeagents into harness-adapter RL to block reward hacking](https://newruntime.com/posts/hugging-face-wires-safeagents-into-harness-adapter-rl-97c5dd.md)
- 2026-11-12 [DeepSeek Harness v0.2.2-rc.2 adds git worktree review and named nested sub-agents](https://newruntime.com/posts/deepseek-harness-v0-2-2-rc-2-adds-30c8a3.md)
- 2026-11-12 [How DXC Technology built a multi-model AI substrate on Amazon Bedrock](https://newruntime.com/posts/how-dxc-technology-built-a-multi-model-ai-ecf016.md)
- 2026-11-12 [Mid-Harness, the source of the TerminalBench-Lite result, changes how the harness selects actions.](https://newruntime.com/posts/mid-harness-the-source-of-the-terminalbench-lite-228a6b.md)
- 2026-11-11 [Building an Automated Improvement Harness for an AI Support Agent Engineering Gen AI LLM ML Infrastructure](https://newruntime.com/posts/building-an-automated-improvement-harness-for-an-ai-74a2f3.md)

## Retrieval (6)

- 2026-11-12 [Exa Search, now in the EU](https://newruntime.com/posts/exa-search-now-in-the-eu-651b57.md)
- 2026-11-12 [Faster Retrieval for Better, More Efficient RAG - Apple Machine Learning Research](https://newruntime.com/posts/faster-retrieval-for-better-more-efficient-rag-apple-e9345b.md)
- 2026-11-12 [Meta AI published model weights on Hugging Face: meta-encoder-v1](https://newruntime.com/posts/meta-ai-published-model-weights-on-hugging-face-2c8206.md)
- 2026-11-10 [Exa launched Exa for Finance, 2B+ financial documents searchable through the existing API](https://newruntime.com/posts/exa-launched-exa-for-finance-2b-financial-documents-63edf8.md)
- 2026-11-10 [Single-shot and agentic retrieval with Amazon Bedrock Managed Knowledge Base](https://newruntime.com/posts/single-shot-and-agentic-retrieval-with-amazon-bedrock-948f68.md)
- 2026-11-09 [Parallel for Legal: web search for legal AI](https://newruntime.com/posts/parallel-for-legal-web-search-for-legal-ai-86a584.md)

## AI adoption (5)

- 2026-11-14 [Microsoft reports post-launch Copilot Cowork adoption inside the company](https://newruntime.com/posts/microsoft-reports-post-launch-copilot-cowork-adoption-inside-9e216e.md)
- 2026-11-11 [Biologists now acknowledge using LLMs almost as much as specialized AI tools like AlphaFold.](https://newruntime.com/posts/biologists-now-acknowledge-using-llms-almost-as-much-4bf81c.md)
- 2026-11-10 [AI Product Management at Spotify: A View From Three Angles](https://newruntime.com/posts/ai-product-management-at-spotify-a-view-from-6017c5.md)
- 2026-11-10 [Helping caregivers use AI in everyday life](https://newruntime.com/posts/helping-caregivers-use-ai-in-everyday-life-859f91.md)
- 2026-11-10 [Work at the Frontier: AI’s effect on freelance work](https://newruntime.com/posts/work-at-the-frontier-ai-s-effect-on-84be35.md)

## Automation (5)

- 2026-11-11 [Atlas On a Roll](https://newruntime.com/posts/atlas-on-a-roll-c3da0b.md)
- 2026-11-11 [Robotics firm Path raises $300 million Series E, $2.45 billion valuation](https://newruntime.com/posts/robotics-firm-path-raises-300-million-series-e-baee99.md)
- 2026-11-10 [Agile Robots Launches Agile ONE SN Humanoid at NVIDIA GTC Berlin and Pilots Humanoids with Audi](https://newruntime.com/posts/agile-robots-launches-agile-one-sn-humanoid-at-51059d.md)
- 2026-11-09 [Automated machine-learning research is becoming a competitive field.](https://newruntime.com/posts/automated-machine-learning-research-is-becoming-a-competitive-a42a5c.md)
- 2026-11-09 [How an assessor could investigate evidence of acceleration from the automation of AI R&D](https://newruntime.com/posts/how-an-assessor-could-investigate-evidence-of-acceleration-47c124.md)

## Developer tools (5)

- 2026-11-13 [Every build machine on Vercel now gets a 5 GB build cache.](https://newruntime.com/posts/every-build-machine-on-vercel-now-gets-a-31e2e3.md)
- 2026-11-12 [AWS has an official local emulator for testing against local clones of its services](https://newruntime.com/posts/aws-has-an-official-local-emulator-for-testing-95fa93.md)
- 2026-11-12 [Zed 1.26.1 preview release adds expandable debugger hovers](https://newruntime.com/posts/zed-1-26-1-preview-release-adds-expandable-a09ecc.md)
- 2026-11-11 [LayerFS 1.1.0 splits the SDK into three packages, adds a CLI, and supports macOS and Windows](https://newruntime.com/posts/layerfs-1-1-0-splits-the-sdk-into-0b1f91.md)
- 2026-11-10 [Extending LLDB to Debug Applications That Span CPUs and GPUs](https://newruntime.com/posts/extending-lldb-to-debug-applications-that-span-cpus-b18d97.md)

## Enterprise AI (5)

- 2026-11-15 [2026: The State of Enterprise AI](https://newruntime.com/posts/2026-the-state-of-enterprise-ai-3fd2e7.md)
- 2026-11-13 [Cohere Expands Sovereign AI Platform with AISLE and cubic](https://newruntime.com/posts/cohere-expands-sovereign-ai-platform-with-aisle-and-421aa7.md)
- 2026-11-11 [AIM Intelligence Presents Enterprise AI Agent Control with NVIDIA at AI Day Seoul](https://newruntime.com/posts/aim-intelligence-presents-enterprise-ai-agent-control-with-605a66.md)
- 2026-11-10 [Introducing Businesses in ChatGPT](https://newruntime.com/posts/introducing-businesses-in-chatgpt-c64a8c.md)
- 2026-11-09 [ChatGPT Enterprise now available in AWS Marketplace](https://newruntime.com/posts/chatgpt-enterprise-now-available-in-aws-marketplace-620566.md)

## Infrastructure (5)

- 2026-11-12 [Uber has open-sourced Shard Manager](https://newruntime.com/posts/uber-has-open-sourced-shard-manager-0d4aaf.md)
- 2026-11-11 [Cloudflare is building an experimental consensus service on the QuePaxa algorithm](https://newruntime.com/posts/cloudflare-is-building-an-experimental-consensus-service-on-904e6a.md)
- 2026-11-11 [Uber says Velox now executes roughly 90% of eligible Presto traffic, processing about 100 petabytes daily](https://newruntime.com/posts/uber-says-velox-now-executes-roughly-90-of-a6efc7.md)
- 2026-11-09 [Every build now gets 64 GB of disk space included for free.](https://newruntime.com/posts/every-build-now-gets-64-gb-of-disk-0fc850.md)
- 2026-11-09 [satmux is an open-source Python tool that routes traffic over pre-warmed TCP connections to a small Linux VM](https://newruntime.com/posts/satmux-is-an-open-source-python-tool-that-879c02.md)

## Memory (5)

- 2026-11-13 [NVLink-sharded Engram tables: Each prefill instance previously stored two 94.4 GiB n-gram tables in host memory, consuming 378 GiB per tray.](https://newruntime.com/posts/nvlink-sharded-engram-tables-each-prefill-instance-previously-907257.md)
- 2026-11-12 [AMD Acquires MEXT to Advance Memory Optimization](https://newruntime.com/posts/amd-acquires-mext-to-advance-memory-optimization-511edd.md)
- 2026-11-12 [SGLang ships Unified Memory, a single GPU arena shared by full-attention KV, sliding-window KV, and recurrent state.](https://newruntime.com/posts/sglang-ships-unified-memory-a-single-gpu-arena-a96b4f.md)
- 2026-11-11 [T-Tech LIMe trains LLMs with about 15% fewer FLOPs without worse quality](https://newruntime.com/posts/t-tech-lime-trains-llms-with-about-15-3b8c2b.md)
- 2026-11-09 [Not Your Diary, Just Your Topic: Grading the Wrong-User Control](https://newruntime.com/posts/not-your-diary-just-your-topic-grading-the-f89387.md)

## Open models (5)

- 2026-11-11 [Baseten joins the NVIDIA Nemotron Coalition to advance custom AI in production](https://newruntime.com/posts/baseten-joins-the-nvidia-nemotron-coalition-to-advance-2a524b.md)
- 2026-11-11 [JEV-35B is an open-weights decision-only model: no chat, just calibrated probabilities over up to 256 options.](https://newruntime.com/posts/jev-35b-is-an-open-weights-decision-only-62c0e9.md)
- 2026-11-11 [Unsloth also provides an NVFP4 build for Nvidia Blackwell GPUs.](https://newruntime.com/posts/unsloth-also-provides-an-nvfp4-build-for-nvidia-d9d35d.md)
- 2026-11-09 [NVIDIA Dynamo Recipes Bring Korea’s Leading Open Models to Production](https://newruntime.com/posts/nvidia-dynamo-recipes-bring-korea-s-leading-open-e4d96c.md)
- 2026-11-09 [Open October LLM Architecture Notes](https://newruntime.com/posts/open-october-llm-architecture-notes-e9c0a6.md)

## Agent security (4)

- 2026-11-14 [OpenAI launches Daybreak Red Team and Daybreak Defense Factory](https://newruntime.com/posts/openai-launches-daybreak-red-team-and-daybreak-defense-97701d.md)
- 2026-11-11 [Malbear Labs details Operation Snowball: jailbroken DeepSeek Flash agent ran a two-week telecom intrusion](https://newruntime.com/posts/malbear-labs-details-operation-snowball-jailbroken-deepseek-flash-ab0725.md)
- 2026-11-11 [NIST publishes draft SP 2400-102e2026 with a taxonomy of attacks on agentic AI systems](https://newruntime.com/posts/nist-publishes-draft-sp-2400-102e2026-with-a-ea025e.md)
- 2026-11-10 [Strengthening Australia against AI-driven cyber incidents: Prime Minister of Australia](https://newruntime.com/posts/strengthening-australia-against-ai-driven-cyber-incidents-prime-f6cc82.md)

## AI (4)

- 2026-11-14 [Approach is training-free: the forward model stays physics-based, autograd only handles gradients.](https://newruntime.com/posts/approach-is-training-free-the-forward-model-stays-81caa3.md)
- 2026-11-11 [Creating a New Internet Bargain to Preserve Fair Use and Protect the Commons](https://newruntime.com/posts/creating-a-new-internet-bargain-to-preserve-fair-1fdeee.md)
- 2026-11-10 [Closing the AI guidance gap in education](https://newruntime.com/posts/closing-the-ai-guidance-gap-in-education-b427e2.md)
- 2026-11-10 [Meta expands Instagram AI age-estimation worldwide to place more teens in Teen Accounts](https://newruntime.com/posts/meta-expands-instagram-ai-age-estimation-worldwide-to-f4fbfd.md)

## AI business models (4)

- 2026-11-12 [Mercor raises $750 million Series D at a $20 billion valuation](https://newruntime.com/posts/mercor-raises-750-million-series-d-at-a-6602aa.md)
- 2026-11-11 [Businesses in ChatGPT: Claim Your Business](https://newruntime.com/posts/businesses-in-chatgpt-claim-your-business-cf67b3.md)
- 2026-11-11 [Chapter 11 Update and Transaction Notice](https://newruntime.com/posts/chapter-11-update-and-transaction-notice-6faab8.md)
- 2026-11-11 [Microsoft recorded $10.0 billion of OpenAI-related revenue in the September 2026 quarter](https://newruntime.com/posts/microsoft-recorded-10-0-billion-of-openai-related-c055b1.md)

## Computer use (4)

- 2026-11-14 [Browser Use shipped a 1Password integration where the agent never sees the password.](https://newruntime.com/posts/browser-use-shipped-a-1password-integration-where-the-5b69de.md)
- 2026-11-14 [Greptile adds computer use to TREX for end-to-end pull-request tests](https://newruntime.com/posts/greptile-adds-computer-use-to-trex-for-end-146673.md)
- 2026-11-14 [Independent bench finds Mage-VL-Cowork asks for screenshots and never clicks outside Microsoft's harness](https://newruntime.com/posts/independent-bench-finds-mage-vl-cowork-asks-for-4341b9.md)
- 2026-11-13 [Hark released Handoff², its second-gen computer-use agent, 2x faster than v1 end-to-end.](https://newruntime.com/posts/hark-released-handoff2-its-second-gen-computer-use-71a7e9.md)

## Local models (4)

- 2026-11-14 [Unsloth released dynamic GGUF quants of Qwen3-4B-Instruct-2507, now at 838K downloads.](https://newruntime.com/posts/unsloth-released-dynamic-gguf-quants-of-qwen3-4b-99c55a.md)
- 2026-11-14 [webAI’s November 9, 2026 v3.1.0 release adds a persistent canvas](https://newruntime.com/posts/webai-s-november-9-2026-v3-1-0-b8e4db.md)
- 2026-11-13 [llama.cpp v1.0.0: A stronger foundation for the local AI ecosystem](https://newruntime.com/posts/llama-cpp-v1-0-0-a-stronger-foundation-67a259.md)
- 2026-11-09 [Manuel García’s Hugging Face page presents dale-0.8b](https://newruntime.com/posts/manuel-garcia-s-hugging-face-page-presents-dale-8a0293.md)

## Sandbox (4)

- 2026-11-13 [Sandbox resourcing: burst, don't reserve](https://newruntime.com/posts/sandbox-resourcing-burst-don-t-reserve-992ab0.md)
- 2026-11-13 [Vercel Sandbox now starts up to 4x faster.](https://newruntime.com/posts/vercel-sandbox-now-starts-up-to-4x-faster-b2aaf9.md)
- 2026-11-10 [Profero publishes unpatched Grok Bot sandbox flaws: one agent command leads to root on the VM and remote access](https://newruntime.com/posts/profero-publishes-unpatched-grok-bot-sandbox-flaws-one-3aef41.md)
- 2026-11-10 [Sandbox snapshots are now up to 10x faster.](https://newruntime.com/posts/sandbox-snapshots-are-now-up-to-10x-faster-1bd269.md)

## Agent interfaces (3)

- 2026-11-12 [feat(telegram): show widgets as interactive Telegram message widgets …](https://newruntime.com/posts/feat-telegram-show-widgets-as-interactive-telegram-message-25e97e.md)
- 2026-11-12 [ui: A library of behaviors for agents to build the next generation of user interfaces](https://newruntime.com/posts/ui-a-library-of-behaviors-for-agents-to-5222af.md)
- 2026-11-09 [OpenAI’s GPT-Live documentation includes a SIP telephony connection guide](https://newruntime.com/posts/openai-s-gpt-live-documentation-includes-a-sip-99c22d.md)

## Agent ops (3)

- 2026-11-12 [Cognition Terraform provider adds Devin Oncall responder resource with Lite mode](https://newruntime.com/posts/cognition-terraform-provider-adds-devin-oncall-responder-resource-5d8a73.md)
- 2026-11-11 [feat: rename off-policy steps to staleness](https://newruntime.com/posts/feat-rename-off-policy-steps-to-staleness-82779d.md)
- 2026-11-11 [OpenClaw's v2026.11.3 release is a maintenance update focused on recovering from interrupted updates](https://newruntime.com/posts/openclaw-s-v2026-11-3-release-is-a-314218.md)

## Agent skills (3)

- 2026-11-12 [NVIDIA open-sourced SkillEvaluator, an Apache-2.0 framework for evaluating agent skills end to end.](https://newruntime.com/posts/nvidia-open-sourced-skillevaluator-an-apache-2-0-7de78b.md)
- 2026-11-10 [genesis-skills: repository for genesis mission skills for AI coding tools](https://newruntime.com/posts/genesis-skills-repository-for-genesis-mission-skills-for-ed2016.md)
- 2026-11-10 [Let the agents triage. Warp has packaged these five checks into an open-source /merge-risk skill.](https://newruntime.com/posts/let-the-agents-triage-warp-has-packaged-these-7669a2.md)

## AI security (3)

- 2026-11-13 [Cyberattack on Fireworks AI allowed hackers to access Notion customers' chat transcripts](https://newruntime.com/posts/cyberattack-on-fireworks-ai-allowed-hackers-to-access-7624d4.md)
- 2026-11-12 [Three Foreign Nationals Arrested And Charged With Scheme To Use Super Intelligence To Defraud Companies Of $129 Million](https://newruntime.com/posts/three-foreign-nationals-arrested-and-charged-with-scheme-42a724.md)
- 2026-11-09 [Understanding AI from the Inside: The State of Mechanistic Interpretability for AI Safety](https://newruntime.com/posts/understanding-ai-from-the-inside-the-state-of-3bf20f.md)

## API design (3)

- 2026-11-12 [OhMyLama’s API guide documents a web-search endpoint at POST /v1/search](https://newruntime.com/posts/ohmylama-s-api-guide-documents-a-web-search-55a968.md)
- 2026-11-11 [Protected Resource Metadata — OAuth 2.0](https://newruntime.com/posts/protected-resource-metadata-oauth-2-0-4bbb5a.md)
- 2026-11-09 [Introducing Sage2: Extraction and Streaming Decisions](https://newruntime.com/posts/introducing-sage2-extraction-and-streaming-decisions-24753e.md)

## Governance (3)

- 2026-11-13 [NPR Machine Gods airs first interview with fired OpenAI safety researchers](https://newruntime.com/posts/npr-machine-gods-airs-first-interview-with-fired-d2119f.md)
- 2026-11-11 [NY Gov. Hochul launches first-in-the-nation state AI testing program](https://newruntime.com/posts/ny-gov-hochul-launches-first-in-the-nation-dac1b5.md)
- 2026-11-11 [Parents sue OpenAI alleging ChatGPT gave fatal incorrect epilepsy medication advice](https://newruntime.com/posts/parents-sue-openai-alleging-chatgpt-gave-fatal-incorrect-016c6a.md)

## Hermes (3)

- 2026-11-11 [Unsloth released Dynamic V3.1 GGUFs for Qwen3.8-27B with top accuracy at every file size](https://newruntime.com/posts/unsloth-released-dynamic-v3-1-ggufs-for-qwen3-9154a0.md)
- 2026-11-10 [Hermes Cloud agents now open in a browser tab with the same UI as Hermes Desktop.](https://newruntime.com/posts/hermes-cloud-agents-now-open-in-a-browser-d7e864.md)
- 2026-11-10 [OpenAI says ChatGPT Business subscribers can now use Sign in with ChatGPT to access a range of third-party](https://newruntime.com/posts/openai-says-chatgpt-business-subscribers-can-now-use-6db52a.md)

## Pricing (3)

- 2026-11-13 [Tokenomics at scale: Track and cap Amazon Bedrock spend per IAM principal with CUR 2.0 and AWS Budgets](https://newruntime.com/posts/tokenomics-at-scale-track-and-cap-amazon-bedrock-2540c2.md)
- 2026-11-11 [Introducing usage-based pricing for Amazon SageMaker AI bidirectional streaming](https://newruntime.com/posts/introducing-usage-based-pricing-for-amazon-sagemaker-ai-8474c0.md)
- 2026-11-11 [MiniMax H3 Image Highspeed API & Pricing](https://newruntime.com/posts/minimax-h3-image-highspeed-api-pricing-0817a6.md)

## Verification (3)

- 2026-11-13 [Math Inc released an executable Lean model of Linux's WireGuard handshake, derived from the C source.](https://newruntime.com/posts/math-inc-released-an-executable-lean-model-of-e56dad.md)
- 2026-11-12 [Semantic diffing of formal policies with anti-unification](https://newruntime.com/posts/semantic-diffing-of-formal-policies-with-anti-unification-907fa6.md)
- 2026-11-12 [Using mathematics to validate network configurations](https://newruntime.com/posts/using-mathematics-to-validate-network-configurations-7d9d32.md)

## Agent runtime (2)

- 2026-11-14 [Architecture patterns for fast session startup on Amazon Bedrock AgentCore runtime](https://newruntime.com/posts/architecture-patterns-for-fast-session-startup-on-amazon-f35fee.md)
- 2026-11-11 [NVIDIA contributes OpenShell agent runtime to CNCF as a Sandbox project](https://newruntime.com/posts/nvidia-contributes-openshell-agent-runtime-to-cncf-as-10b351.md)

## Enterprise agents (2)

- 2026-11-13 [SSO via Okta, Entra ID, or any SAML/OIDC provider now available on the Teams plan.](https://newruntime.com/posts/sso-via-okta-entra-id-or-any-saml-000a61.md)
- 2026-11-12 [Databricks has introduced a new Genie release that expands the product from conversational analytics into a long-running enterprise agent.](https://newruntime.com/posts/databricks-has-introduced-a-new-genie-release-that-48a97e.md)

## Future of work (2)

- 2026-11-13 [State of the tech job market: late 2026](https://newruntime.com/posts/state-of-the-tech-job-market-late-2026-922c49.md)
- 2026-11-13 [The AI Layoff Trap](https://newruntime.com/posts/the-ai-layoff-trap-a8330e.md)

## Interfaces (2)

- 2026-11-12 [Mario 64 mixed-reality mod runs at 90fps](https://newruntime.com/posts/mario-64-mixed-reality-mod-runs-at-90fps-dd6fe4.md)
- 2026-11-11 [Retiring Advanced and Standard Voice in ChatGPT](https://newruntime.com/posts/retiring-advanced-and-standard-voice-in-chatgpt-2ae063.md)

## Model routing (2)

- 2026-11-13 [Parallel Decisions now available on AI Gateway for free](https://newruntime.com/posts/parallel-decisions-now-available-on-ai-gateway-for-9c8990.md)
- 2026-11-11 [Introducing Baseten Router: a Pareto-optimal model router for coding tasks](https://newruntime.com/posts/introducing-baseten-router-a-pareto-optimal-model-router-db4bc1.md)

## Multi-agent (2)

- 2026-11-12 [Paper shows activation probes for multi-agent LLM collusion mostly detect the instruction, not the decision](https://newruntime.com/posts/paper-shows-activation-probes-for-multi-agent-llm-90905b.md)
- 2026-11-10 [s$^3$: Simple Scaling of Swarms via Cooperative Multi-agent RL](https://newruntime.com/posts/s-3-simple-scaling-of-swarms-via-cooperative-a70e2e.md)

## Observability (2)

- 2026-11-10 [Amazon Bedrock tokenomics data lake](https://newruntime.com/posts/amazon-bedrock-tokenomics-data-lake-442b50.md)
- 2026-11-10 [Speed Insights and Web Analytics now support custom environments.](https://newruntime.com/posts/speed-insights-and-web-analytics-now-support-custom-9580f4.md)

## Prompts (2)

- 2026-11-10 [New projection-free evolution strategy for black-box prompt search beats BBT on GLUE at equal cost.](https://newruntime.com/posts/new-projection-free-evolution-strategy-for-black-box-575113.md)
- 2026-11-09 [Follow-up GazeQwen achieves 63.9% by modulating the LLM internally rather than via prompts.](https://newruntime.com/posts/follow-up-gazeqwen-achieves-63-9-by-modulating-526137.md)

## Workflows (2)

- 2026-11-13 [Workflow SDK 5 is now available](https://newruntime.com/posts/workflow-sdk-5-is-now-available-0adb56.md)
- 2026-11-11 [Modal Labs ships modal-dojo 0.1.1 with new recipes and default proxy-token auth](https://newruntime.com/posts/modal-labs-ships-modal-dojo-0-1-1-611eca.md)

## Agent economics (1)

- 2026-11-13 [Optimizing cost per task and proving agent ROI on Amazon Bedrock](https://newruntime.com/posts/optimizing-cost-per-task-and-proving-agent-roi-b45b0e.md)

## Agent identity (1)

- 2026-11-11 [Linux Foundation takes on Agent Name Service, a DNS-based open standard for AI agent identity](https://newruntime.com/posts/linux-foundation-takes-on-agent-name-service-a-b6ccb4.md)

## Agent loops (1)

- 2026-11-12 [Introducing the Web Search API on Amazon Bedrock AgentCore](https://newruntime.com/posts/introducing-the-web-search-api-on-amazon-bedrock-c1b723.md)

## Agent observability (1)

- 2026-11-10 [Langfuse October Update - Langfuse](https://newruntime.com/posts/langfuse-october-update-langfuse-cbea97.md)

## Agent protocols (1)

- 2026-11-13 [Sierra releases Personal Agent Protocol (Poppy) 1.0, an open standard for personal agents to interact with companies](https://newruntime.com/posts/sierra-releases-personal-agent-protocol-poppy-1-0-dd15ec.md)

## Agent tools (1)

- 2026-11-10 [Patch: Build and play](https://newruntime.com/posts/patch-build-and-play-8af3e7.md)

## Design systems (1)

- 2026-11-14 [OpenAI’s Apps SDK UI is a React component library and design system for building interfaces in ChatGPT apps](https://newruntime.com/posts/openai-s-apps-sdk-ui-is-a-react-da7849.md)

## Frontend (1)

- 2026-11-13 [Open-source chat-demo frontend ships as reference implementation with import/export support.](https://newruntime.com/posts/open-source-chat-demo-frontend-ships-as-reference-f2c95c.md)

## Generative UI (1)

- 2026-11-09 [Chat SDK agents can now stream generative UI with json-render](https://newruntime.com/posts/chat-sdk-agents-can-now-stream-generative-ui-e167d7.md)

## Knowledge work (1)

- 2026-11-14 [Six Surge Benchmarks Join the Fireworks Specialized Intelligence Index](https://newruntime.com/posts/six-surge-benchmarks-join-the-fireworks-specialized-intelligence-38d69e.md)

## Orchestration (1)

- 2026-11-10 [How We Built TLX: A Hardware-Native GPU Compiler for Production ML at Scale](https://newruntime.com/posts/how-we-built-tlx-a-hardware-native-gpu-31371f.md)

## Personalization (1)

- 2026-11-10 [Your Algorithm: Giving People More Ways to Personalize Facebook Recommendations](https://newruntime.com/posts/your-algorithm-giving-people-more-ways-to-personalize-ea5262.md)

## rumors (1)

- 2026-11-10 [Unconfirmed: Grok Bot weekly limits reportedly reset for all users](https://newruntime.com/posts/unconfirmed-grok-bot-weekly-limits-reportedly-reset-for-all-38ad9f.md)

## Tool use (1)

- 2026-11-11 [BotPsy: On-Device Personal Assistant with a Sub-Billion-Parameter Agentic Language Model](https://newruntime.com/posts/botpsy-on-device-personal-assistant-with-a-sub-92ce02.md)

## Other notes (22)

- 2026-11-15 [AWS makes next-generation AgentCore runtime the default for new Bedrock AgentCore runtimes](https://newruntime.com/posts/aws-makes-next-generation-agentcore-runtime-the-default-869851.md)
- 2026-11-15 [Ted Tschopp’s essay proposes four checks before a company approves a purchase recommended by an AI assistant](https://newruntime.com/posts/ted-tschopp-s-essay-proposes-four-checks-before-4cb0d5.md)
- 2026-11-14 [Agile Robots raises $500 million Series F at an $8.5 billion valuation, led by SoftBank](https://newruntime.com/posts/agile-robots-raises-500-million-series-f-at-7ac7cd.md)
- 2026-11-14 [Agility Robotics is going public through a SPAC transaction that TechCrunch reports will raise $465 million](https://newruntime.com/posts/agility-robotics-is-going-public-through-a-spac-7c55c2.md)
- 2026-11-14 [browser-use 1.0.4 batches screenshot pruning and halves BU Bench model cost](https://newruntime.com/posts/browser-use-1-0-4-batches-screenshot-pruning-3b3893.md)
- 2026-11-14 [Goodwin Advises Foundry Robotics on its $100 Million Series A Financing and Acquisition of Avea Robotics](https://newruntime.com/posts/goodwin-advises-foundry-robotics-on-its-100-million-1374c0.md)
- 2026-11-14 [Labels oppose Suno's full stay in UMG II, propose a partial stay with targeted discovery on Suno v6](https://newruntime.com/posts/labels-oppose-suno-s-full-stay-in-umg-b6804f.md)
- 2026-11-14 [Nominal raises $240M at a $3B valuation to build the intelligence platform for hardware engineering](https://newruntime.com/posts/nominal-raises-240m-at-a-3b-valuation-to-beec10.md)
- 2026-11-14 [OpenCode Go Teams now requires two members, a $40 monthly floor](https://newruntime.com/posts/opencode-go-teams-now-requires-two-members-a-4e61bb.md)
- 2026-11-14 [Silicium Rose disputes Scantist’s “GLM-5.3-class” claim for the 192-expert GGUF](https://newruntime.com/posts/silicium-rose-disputes-scantist-s-glm-5-3-69da24.md)
- 2026-11-14 [Snorkel finds Fable 5.2 pass@1 at 71% versus 53% for Fable 5.1 on matched coding tasks](https://newruntime.com/posts/snorkel-finds-fable-5-2-pass-1-at-34944d.md)
- 2026-11-14 [TIL: Drop Columns a Release Later](https://newruntime.com/posts/til-drop-columns-a-release-later-6e1ee7.md)
- 2026-11-13 [Findings of Fact, Conclusions of Law and Judgment](https://newruntime.com/posts/findings-of-fact-conclusions-of-law-and-judgment-ed6fb0.md)
- 2026-11-12 [A group maintaining a public Yang–Mills results repository says its assessment of OpenAI’s 28 October 2026](https://newruntime.com/posts/a-group-maintaining-a-public-yang-mills-results-0e95d8.md)
- 2026-11-12 [Apple and Microsoft added to Russia’s information-distributors register](https://newruntime.com/posts/apple-and-microsoft-added-to-russia-s-information-a212a9.md)
- 2026-11-12 [Court orders Abdulrahman Alfozan detained pending trial](https://newruntime.com/posts/court-orders-abdulrahman-alfozan-detained-pending-trial-638f4d.md)
- 2026-11-12 [NVIDIA Announces CEO Transition](https://newruntime.com/posts/nvidia-announces-ceo-transition-b4bced.md)
- 2026-11-11 [Court sets July 19, 2027 trial in Apple’s trade-secrets suit against OpenAI](https://newruntime.com/posts/court-sets-july-19-2027-trial-in-apple-e8a5e4.md)
- 2026-11-11 [Knowledge graphs drift out of date as organizations merge, people change jobs, new entities appear, and old relationships expire.](https://newruntime.com/posts/knowledge-graphs-drift-out-of-date-as-organizations-b8e6b6.md)
- 2026-11-11 [Meta FAIR releases e-means, a mini-batch k-means variant now in Faiss 1.16.0](https://newruntime.com/posts/meta-fair-releases-e-means-a-mini-batch-b88694.md)
- 2026-11-11 [SINGULUS TECHNOLOGIES says it has signed a contract with SpaceX for a turnkey line to manufacture](https://newruntime.com/posts/singulus-technologies-says-it-has-signed-a-contract-b63b4b.md)
- 2026-11-09 [Nvidia’s board has opened an independent investigation into its own CEO.](https://newruntime.com/posts/nvidia-s-board-has-opened-an-independent-investigation-fa507a.md)

## Other weeks

- Previous: [Nov 2–8, 2026](https://newruntime.com/weeks/2026-w45.md)
- All weeks: https://newruntime.com/weeks.md