New Runtime data
Noticed in the data
Changes New Runtime's collectors detected automatically in public data: a hiring surge, a pricing page edit, an adoption spike. Each links to the public page where it can be seen. A change is a signal, not a story, until a primary source explains it. Updated 2026-10-28.
150 changesCompanies in the dataJSON
Google DeepMind Google: Gemma 4 26B A4B : list price changed (OpenRouter)
Per million tokens: input $0.0765 → $0.09, output $0.255 → $0.3, cache_read $0.0425 → $0.05. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
NVIDIA NVIDIA: Nemotron 3 Super: list price changed (OpenRouter)
Per million tokens: input $0.08 → $0.3, output $0.45 → $2.4. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
xAI posted 2 new data centers roles
New on xAI's job board since the last check: Structural Engineer, Data Center - Memphis; Supervisor, Data Center Operations - Memphis. Roles in data centers point at work beyond the company's current products; one posting is a hint, several are a plan.
DeepSeek DeepSeek: DeepSeek V4 Pro 0423: list price changed (OpenRouter)
Per million tokens: input $0.8754 → $0.9553, output $1.7508 → $1.9105, cache_read $0.0729 → $0.0796. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
OpenRouter policy page: policy text changed
1 lines added, 2 removed on the policy page; an observed change, not yet a confirmed announcement.
OpenRouter models page: new model listed
Now listed: ternary-bonsai-2-9b. Limits before: 0.0000000701 context, 0.00000008 context, 0.00000009464 context, 0.0000001694 context, 0.00000033 context, 0.0000006 context; after: 0.00000004 context, 0.0000000798 context, 0.00000013 context, 0.0000001568 context, 0.00000028 context, 0.00000045 context. 20 lines added, 19 removed on the models page; an observed change, not yet a confirmed announcement.
Perplexity pricing page: prices changed
Prices before: none; after: $0.001, $0.005, $1.00, $5.00, $50.00. 30 lines added, 2 removed on the pricing page; an observed change, not yet a confirmed announcement.
Vercel pricing page: page content changed
1 lines added, 1 removed on the pricing page; an observed change, not yet a confirmed announcement.
Google DeepMind pricing page: page content changed
2 lines added, 2 removed on the pricing page; an observed change, not yet a confirmed announcement.
Together AI limits page: page content changed
4 lines added, 9 removed on the limits page; an observed change, not yet a confirmed announcement.
Microsoft changelog page: changelog updated
Now listed: oct.27. 3 lines added, 0 removed on the changelog page; an observed change, not yet a confirmed announcement.
OpenAI changelog page: changelog updated
Now listed: alpha.1.2, rust-v0.165.0. 102 lines added, 104 removed on the changelog page; an observed change, not yet a confirmed announcement.
Vercel changelog page: changelog updated
9 lines added, 9 removed on the changelog page; an observed change, not yet a confirmed announcement.
ElevenLabs models page: model no longer listed
No longer listed: eleven-v3, eleven-v3-conversational, eleven-v4, eleven-v4-turbo, flash-v25, multilingual-v2, scribe-v2, scribe-v2-medical. 825 lines added, 386 removed on the models page; an observed change, not yet a confirmed announcement.
Vercel models page: prices changed
Prices before: $0.02; after: $0.07. 9 lines added, 9 removed on the models page; an observed change, not yet a confirmed announcement.
Anthropic pricing page: page content changed
0 lines added, 1 removed on the pricing page; an observed change, not yet a confirmed announcement.
Anthropic changelog page: changelog updated
41 lines added, 37 removed on the changelog page; an observed change, not yet a confirmed announcement.
Cohere changelog page: changelog updated
No longer listed: c4ai-command-a-03-2025, c4ai-command-r-plus, c4ai-command-r7b-12-2024, cohere-embed-v5-fast, cohere-embed-v5-pro, cohere-parse-v5, cohere-rerank-v5-fast, cohere-rerank-v5-pro. Limits before: 128,000 tokens, 128k context, 128k tokens, 256k context, 256k tokens, 3. context; after: none. 289 lines added, 448 removed on the changelog page; an observed change, not yet a confirmed announcement.
Cohere limits page: new model listed
Now listed: command-r-0824. 318 lines added, 38 removed on the limits page; an observed change, not yet a confirmed announcement.
Cohere models page: model no longer listed
Now listed: command-r-0824. No longer listed: command-r7b, rerank-1. 817 lines added, 149 removed on the models page; an observed change, not yet a confirmed announcement.
Cohere deprecations page: new model on the deprecations page
Now listed: command-r-0824. 348 lines added, 64 removed on the deprecations page; an observed change, not yet a confirmed announcement.
Moonshot AI MoonshotAI: Kimi K3: list price changed (OpenRouter)
Per million tokens: input $0.159 → $0.179, cache_read $0.149 → $0.169. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Vercel policy page: policy text changed
62 lines added, 62 removed on the policy page; an observed change, not yet a confirmed announcement.
Together AI models page: prices changed
Prices before: $0.0075, $0.075, $0.1625, $0.325, $0.375, $1.50, $1.65, $1.75; after: $0.01875, $0.1875, $0.75. Limits before: 1m tokens; after: none. 9 lines added, 42 removed on the models page; an observed change, not yet a confirmed announcement.
OpenRouter pricing page: page content changed
1 lines added, 2 removed on the pricing page; an observed change, not yet a confirmed announcement.
ElevenLabs changelog page: changelog updated
No longer listed: eleven-v4, gemini-3.6-flash, gemini-3.7-flash, gpt-5.6-luna, gpt-5.6-sol, gpt-5.6-terra, list-v-2, scribe-v2-medical. 2314 lines added, 1231 removed on the changelog page; an observed change, not yet a confirmed announcement.
OpenRouter changelog page: changelog updated
4 lines added, 4 removed on the changelog page; an observed change, not yet a confirmed announcement.
Moonshot AI MoonshotAI: Kimi K3: list price changed (OpenRouter)
Per million tokens: input $0.189 → $0.159, output $20 → $11.5, cache_read $0.179 → $0.149. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Alibaba Qwen Qwen: Qwen3.6 35B A3B: list price changed (OpenRouter)
Per million tokens: cache_read $0.05 → $0.025. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
DeepSeek DeepSeek: DeepSeek V4 Pro 0813: list price changed (OpenRouter)
Per million tokens: input $0.66 → $1.32, output $1.98 → $3.96, cache_read $0.022 → $0.044. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
DeepSeek DeepSeek: DeepSeek V4.1 Flash 1017: list price changed (OpenRouter)
Per million tokens: input $0.15 → $0.3, output $0.6 → $1.2, cache_read $0.003 → $0.006. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Moonshot AI MoonshotAI: Kimi K3: list price changed (OpenRouter)
Per million tokens: input $0.15 → $0.189, cache_read $0.14 → $0.179. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Alibaba Qwen Qwen: Qwen3.6 27B: list price changed (OpenRouter)
Per million tokens: input $0.32 → $0.45, output $3.2 → $2.7. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Alibaba Qwen Qwen: Qwen3.8 27B: list price changed (OpenRouter)
Per million tokens: input $0.2 → $0.039, output $2.5 → $1.25, cache_read $0.09 → $0.0039. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Meta AI Meta: Llama 3.1 8B Instruct: list price changed (OpenRouter)
Per million tokens: input $0.05 → $0.02, output $0.08 → $0.04. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
NVIDIA NVIDIA: Nemotron 3 Super: list price changed (OpenRouter)
Per million tokens: input $0.3 → $0.08, output $2.4 → $0.45. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
DeepSeek DeepSeek: DeepSeek V4 Flash 0423: list price changed (OpenRouter)
Per million tokens: input $0.0473 → $0.14, output $0.0946 → $0.28, cache_read $0.0095 → $0.028. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
reason-machines/reason-network: 161 commits last week, manyx its usual pace
reason-machines/reason-network had 161 commits in the last complete week against a median of 0 before. A burst of work often precedes a release.
eagle-hq/EAGLE-4 shipped 4 releases in four weeks, faster than its usual pace
eagle-hq/EAGLE-4 published 4 releases in the last 28 days against 0 in the 56 days before.
Tencent-Hunyuan/Hy-Memory shipped 4 releases in four weeks, faster than its usual pace
Tencent-Hunyuan/Hy-Memory published 4 releases in the last 28 days against 0 in the 56 days before.
chenglou/handwired: 508 commits last week, 19.2x its usual pace
chenglou/handwired had 508 commits in the last complete week against a median of 27 before. A burst of work often precedes a release.
trycua/cua shipped 60 releases in four weeks, faster than its usual pace
trycua/cua published 60 releases in the last 28 days against 0 in the 56 days before.
CRTnoise/ember-p2p shipped 7 releases in four weeks, faster than its usual pace
CRTnoise/ember-p2p published 7 releases in the last 28 days against 0 in the 56 days before.
cobanov/PS5LM shipped 4 releases in four weeks, faster than its usual pace
cobanov/PS5LM published 4 releases in the last 28 days against 0 in the 56 days before.
openqodex/openqodex shipped 20 releases in four weeks, faster than its usual pace
openqodex/openqodex published 20 releases in the last 28 days against 0 in the 56 days before.
init-co-jp/diff-tour shipped 6 releases in four weeks, faster than its usual pace
init-co-jp/diff-tour published 6 releases in the last 28 days against 0 in the 56 days before.
anthropics/claude-plugins-official: 214 commits last week, 4x its usual pace
anthropics/claude-plugins-official had 214 commits in the last complete week against a median of 54 before. A burst of work often precedes a release.
storytold/effectcraft shipped 17 releases in four weeks, faster than its usual pace
storytold/effectcraft published 17 releases in the last 28 days against 0 in the 56 days before.
storytold/cadcraft: 111 commits last week, 11.7x its usual pace
storytold/cadcraft had 111 commits in the last complete week against a median of 10 before. A burst of work often precedes a release.
storytold/cadcraft shipped 11 releases in four weeks, faster than its usual pace
storytold/cadcraft published 11 releases in the last 28 days against 0 in the 56 days before.
storytold/filmcraft shipped 14 releases in four weeks, faster than its usual pace
storytold/filmcraft published 14 releases in the last 28 days against 0 in the 56 days before.
1homsi/thehumanbox shipped 60 releases in four weeks, faster than its usual pace
1homsi/thehumanbox published 60 releases in the last 28 days against 0 in the 56 days before.
reason-machines/reason-network shipped 26 releases in four weeks, faster than its usual pace
reason-machines/reason-network published 26 releases in the last 28 days against 0 in the 56 days before.
simbasjungles/free-my-tv: 112 commits last week, manyx its usual pace
simbasjungles/free-my-tv had 112 commits in the last complete week against a median of 0 before. A burst of work often precedes a release.
BerriAI/litellm shipped 59 releases in four weeks, faster than its usual pace
BerriAI/litellm published 59 releases in the last 28 days against 1 in the 56 days before.
hardenedlinux/SonicBoom: 216 commits last week, 12x its usual pace
hardenedlinux/SonicBoom had 216 commits in the last complete week against a median of 18 before. A burst of work often precedes a release.
meshy-dev/meshy-mcp-server shipped 7 releases in four weeks, faster than its usual pace
meshy-dev/meshy-mcp-server published 7 releases in the last 28 days against 1 in the 56 days before.
Gathin23/tv-vibehack: 831 commits last week, manyx its usual pace
Gathin23/tv-vibehack had 831 commits in the last complete week against a median of 0 before. A burst of work often precedes a release.
Gathin23/tv-vibehack shipped 9 releases in four weeks, faster than its usual pace
Gathin23/tv-vibehack published 9 releases in the last 28 days against 0 in the 56 days before.
YuJunZhiXue/dsh-purge shipped 10 releases in four weeks, faster than its usual pace
YuJunZhiXue/dsh-purge published 10 releases in the last 28 days against 0 in the 56 days before.
glanderness/BeefTV: 234 commits last week, 3x its usual pace
glanderness/BeefTV had 234 commits in the last complete week against a median of 77 before. A burst of work often precedes a release.
perplexityai/SPACE: 234 commits last week, 6.3x its usual pace
perplexityai/SPACE had 234 commits in the last complete week against a median of 37 before. A burst of work often precedes a release.
glanderness/BeefTV shipped 43 releases in four weeks, faster than its usual pace
glanderness/BeefTV published 43 releases in the last 28 days against 17 in the 56 days before.
huggingface/OpenEnv shipped 4 releases in four weeks, faster than its usual pace
huggingface/OpenEnv published 4 releases in the last 28 days against 3 in the 56 days before.
NVIDIA/flare-local-inference: 329 commits last week, manyx its usual pace
NVIDIA/flare-local-inference had 329 commits in the last complete week against a median of 0 before. A burst of work often precedes a release.
DeepSeek DeepSeek: DeepSeek V4 Pro 0423: list price changed (OpenRouter)
Per million tokens: input $0.8771 → $0.8754, output $1.7543 → $1.7508, cache_read $0.0731 → $0.0729. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Moonshot AI MoonshotAI: Kimi K3: list price changed (OpenRouter)
Per million tokens: input $0.2 → $0.15, output $11.5 → $20, cache_read $0.19 → $0.14. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
OpenRouter lists z-ai/glm-4.5v for shutdown on 2026-12-31
OpenRouter now lists z-ai/glm-4.5v with a shutdown date of 2026-12-31 (65 days from now). Requests to the model are expected to fail after that date.
OpenRouter lists meta-llama/llama-3.1-70b-instruct for shutdown on 2026-11-03
OpenRouter now lists meta-llama/llama-3.1-70b-instruct with a shutdown date of 2026-11-03 (7 days from now). Requests to the model are expected to fail after that date.
OpenRouter: meta-llama/llama-3.1-70b-instruct shuts down in 7 days (2026-11-03)
OpenRouter shuts down meta-llama/llama-3.1-70b-instruct on 2026-11-03.
OpenAI models page: page content changed
1 lines added, 1 removed on the models page; an observed change, not yet a confirmed announcement.
OpenAI pricing page: page content changed
1 lines added, 1 removed on the pricing page; an observed change, not yet a confirmed announcement.
OpenAI limits page: page content changed
1 lines added, 1 removed on the limits page; an observed change, not yet a confirmed announcement.
OpenAI pricing page: page content changed
1 lines added, 1 removed on the pricing page; an observed change, not yet a confirmed announcement.
Anthropic limits page: page content changed
17 lines added, 17 removed on the limits page; an observed change, not yet a confirmed announcement.
Hugging Face pricing page: page content changed
2 lines added, 2 removed on the pricing page; an observed change, not yet a confirmed announcement.
Cohere pricing page: page content changed
2 lines added, 2 removed on the pricing page; an observed change, not yet a confirmed announcement.
Cohere policy page: policy text changed
2 lines added, 2 removed on the policy page; an observed change, not yet a confirmed announcement.
Cohere changelog page: changelog updated
No longer listed: north-mini-code-1-0, north-small-translate-09-2026. 2 lines added, 2 removed on the changelog page; an observed change, not yet a confirmed announcement.
Cerebras pricing page: page content changed
1 lines added, 0 removed on the pricing page; an observed change, not yet a confirmed announcement.
Anthropic pricing page: page content changed
10 lines added, 10 removed on the pricing page; an observed change, not yet a confirmed announcement.
Anthropic models page: page content changed
165 lines added, 23 removed on the models page; an observed change, not yet a confirmed announcement.
Anthropic limits page: page content changed
3 lines added, 3 removed on the limits page; an observed change, not yet a confirmed announcement.
Runway pricing page: page content changed
9 lines added, 0 removed on the pricing page; an observed change, not yet a confirmed announcement.
Anthropic changelog page: changelog updated
224 lines added, 214 removed on the changelog page; an observed change, not yet a confirmed announcement.
Google DeepMind Google: Gemma 4 26B A4B : list price changed (OpenRouter)
Per million tokens: input $0.09 → $0.0765, output $0.3 → $0.255, cache_read $0.05 → $0.0425. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Vercel models page: model no longer listed
Now listed: ash-v2.5, cribe-v2, e-v1-max, eleven-v3-conver, fects-v2, flash-v2, leven-v3, leven-v4. No longer listed: cury-2.5, deepseek-v4-flash-vi, gemini-3, gemini-3.8-f, gemini-3.8-flash, k-v5-pro, ling-3, marize-1. Prices before: $0.03, $0.04, $0.05, $0.12, $0.15, $0.22, $0.30, $0.32; after: $0.08, $0.11, $0.39, $2.40, $20, $3, $4.50, $40. 62 lines added, 80 removed on the models page; an observed change, not yet a confirmed announcement.
LangChain changelog page: changelog updated
Prices before: none; after: $ 0.00026, $0,, $0.0003. 428 lines added, 0 removed on the changelog page; an observed change, not yet a confirmed announcement.
OpenRouter models page: page content changed
Limits before: 0.0000001151 context, 0.0000001637 context, 0.0000003 context, 0.00000045 context, 0.000000528 context, 0.00000058212 context; after: 0.0000000701 context, 0.00000009464 context, 0.0000001694 context, 0.000000255 context, 0.00000033 context, 0.00000146 context. 12 lines added, 12 removed on the models page; an observed change, not yet a confirmed announcement.
Alibaba Qwen pricing page: new model listed
Now listed: qwen-audio-3.1-realtime-flash. Prices before: none; after: $0.15, $0.212, $0.23, $0.47, $0.636, $0.7, $0.848, $0.93. Limits before: none; after: 1,000,000 tokens. 26 lines added, 0 removed on the pricing page; an observed change, not yet a confirmed announcement.
Anthropic deprecations page: page content changed
2 lines added, 2 removed on the deprecations page; an observed change, not yet a confirmed announcement.
DeepSeek DeepSeek: DeepSeek V4 Pro 0423: list price changed (OpenRouter)
Per million tokens: input $0.8804 → $0.8771, output $1.7609 → $1.7543, cache_read $0.0734 → $0.0731. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
OpenAI changelog page: changelog updated
2 lines added, 2 removed on the changelog page; an observed change, not yet a confirmed announcement.
Vercel changelog page: changelog updated
No longer listed: gpt-6.1. 7 lines added, 6 removed on the changelog page; an observed change, not yet a confirmed announcement.
OpenAI changelog page: changelog updated
1 lines added, 1 removed on the changelog page; an observed change, not yet a confirmed announcement.
OpenAI deprecations page: page content changed
1 lines added, 1 removed on the deprecations page; an observed change, not yet a confirmed announcement.
Google DeepMind pricing page: prices changed
Prices before: $0.1, $0.2, $1, $1.1, $10, $11, $18, $19.8; after: $0.0000005, $0.000001, $0.10, $0.20, $1.00, $1.10, $10.00, $11.00. Limits before: none; after: 1m tokens. 67 lines added, 39 removed on the pricing page; an observed change, not yet a confirmed announcement.
Cohere models page: model no longer listed
No longer listed: north-mini-code-1-0, north-small-translate-09-2026. 3 lines added, 3 removed on the models page; an observed change, not yet a confirmed announcement.
Moonshot AI MoonshotAI: Kimi K2.7 Code: list price changed (OpenRouter)
Per million tokens: input $0.6725 → $0.56. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Moonshot AI MoonshotAI: Kimi K3: list price changed (OpenRouter)
Per million tokens: input $0.15 → $0.2, cache_read $0.14 → $0.19. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
DeepSeek DeepSeek: DeepSeek V4 Flash 0423: list price changed (OpenRouter)
Per million tokens: input $0.0449 → $0.0473, output $0.45 → $0.0946, cache_read $0.0449 → $0.0095. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Moonshot AI MoonshotAI: Kimi K2.7 Code: list price changed (OpenRouter)
Per million tokens: input $0.56 → $0.6725. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
DeepSeek DeepSeek: DeepSeek V4 Pro 0423: list price changed (OpenRouter)
Per million tokens: input $0.8973 → $0.8804, output $1.7946 → $1.7609, cache_read $0.0748 → $0.0734. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
DeepSeek DeepSeek: DeepSeek V4 Flash 0423: list price changed (OpenRouter)
Per million tokens: input $0.05 → $0.0449, cache_read $0.05 → $0.0449. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Alibaba Qwen pricing page: prices changed
Prices before: none; after: $0.014, $0.016, $0.113, $0.15, $0.212, $0.23, $0.382, $0.47. 26 lines added, 26 removed on the pricing page; an observed change, not yet a confirmed announcement.
OpenRouter models page: model no longer listed
No longer listed: nemotron-3-super-120b-a12b:free. Limits before: 0 context, 0.0000000701 context, 0.0000001568 context, 0.00000028 context, 0.0000012 context, 0.00000191052 context; after: 0.0000001151 context, 0.0000001637 context, 0.00000045 context, 0.00000058212 context, 0.0000006 context, 0.00000198 context. 11 lines added, 12 removed on the models page; an observed change, not yet a confirmed announcement.
DeepSeek DeepSeek: DeepSeek V4 Pro 0423: list price changed (OpenRouter)
Per million tokens: input $0.9553 → $0.8973, output $1.9105 → $1.7946, cache_read $0.0796 → $0.0748. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Microsoft pricing page: page content changed
2 lines added, 2 removed on the pricing page; an observed change, not yet a confirmed announcement.
Google DeepMind pricing page: page content changed
2 lines added, 2 removed on the pricing page; an observed change, not yet a confirmed announcement.
Replit changelog page: changelog updated
4 lines added, 3 removed on the changelog page; an observed change, not yet a confirmed announcement.
Vercel models page: prices changed
Prices before: $0.07; after: $0.03. 12 lines added, 13 removed on the models page; an observed change, not yet a confirmed announcement.
Moonshot AI MoonshotAI: Kimi K3: list price changed (OpenRouter)
Per million tokens: input $0.2 → $0.159, output $11.5 → $20, cache_read $0.19 → $0.149. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
DeepSeek DeepSeek: DeepSeek V4.1 Flash 1017: list price changed (OpenRouter)
Per million tokens: input $0.3 → $0.15, output $1.2 → $0.6, cache_read $0.006 → $0.003. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
DeepSeek DeepSeek: DeepSeek V4 Pro 0813: list price changed (OpenRouter)
Per million tokens: input $1.32 → $0.66, output $3.96 → $1.98, cache_read $0.044 → $0.022. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Meta AI Meta: Llama 3.1 70B Instruct is listed to retire on 2026-11-03
OpenRouter now gives Meta: Llama 3.1 70B Instruct a retirement date of 2026-11-03. The vendor's deprecation page is the authority.
Moonshot AI MoonshotAI: Kimi K3: list price changed (OpenRouter)
Per million tokens: output $20 → $11.5. Seen in the OpenRouter listing; the vendor's pricing page is the authority.
Amazon Bedrock: anthropic.claude-3-5-sonnet-20240620-v1:0 shuts down in 17 days (2026-11-13)
Amazon Bedrock shuts down anthropic.claude-3-5-sonnet-20240620-v1:0 on 2026-11-13.
Amazon Bedrock: anthropic.claude-3-5-sonnet-20241022-v2:0 shuts down in 17 days (2026-11-13)
Amazon Bedrock shuts down anthropic.claude-3-5-sonnet-20241022-v2:0 on 2026-11-13.
OpenRouter: poolside/laguna-xs-2.1 shuts down in 4 days (2026-10-31)
OpenRouter shuts down poolside/laguna-xs-2.1 on 2026-10-31.
OpenRouter: nousresearch/hermes-3-llama-3.1-405b shuts down in 6 days (2026-11-02)
OpenRouter shuts down nousresearch/hermes-3-llama-3.1-405b on 2026-11-02.
OpenRouter: nousresearch/hermes-3-llama-3.1-70b shuts down in 6 days (2026-11-02)
OpenRouter shuts down nousresearch/hermes-3-llama-3.1-70b on 2026-11-02.
OpenRouter: upstage/solar-pro-3 shuts down in 3 days (2026-10-30)
OpenRouter shuts down upstage/solar-pro-3 on 2026-10-30.
Amazon Bedrock: anthropic.claude-3-haiku-20240307-v1:0 shuts down in 17 days (2026-11-13)
Amazon Bedrock shuts down anthropic.claude-3-haiku-20240307-v1:0 on 2026-11-13.
Amazon Bedrock: anthropic.claude-3-sonnet-20240229-v1:0 shuts down in 17 days (2026-11-13)
Amazon Bedrock shuts down anthropic.claude-3-sonnet-20240229-v1:0 on 2026-11-13.
OpenRouter: bytedance-seed/seed-1.6 shuts down in 15 days (2026-11-11)
OpenRouter shuts down bytedance-seed/seed-1.6 on 2026-11-11.
OpenRouter: poolside/laguna-s-2.1 shuts down in 4 days (2026-10-31)
OpenRouter shuts down poolside/laguna-s-2.1 on 2026-10-31.
OpenRouter: bytedance-seed/seed-1.6-flash shuts down in 15 days (2026-11-11)
OpenRouter shuts down bytedance-seed/seed-1.6-flash on 2026-11-11.
OpenRouter: bytedance-seed/seed-2.0-code shuts down in 15 days (2026-11-11)
OpenRouter shuts down bytedance-seed/seed-2.0-code on 2026-11-11.
Amazon Bedrock: ai21.jamba-1-5-large-v1:0 shuts down in 30 days (2026-11-26)
Amazon Bedrock shuts down ai21.jamba-1-5-large-v1:0 on 2026-11-26.
Amazon Bedrock: ai21.jamba-1-5-mini-v1:0 shuts down in 30 days (2026-11-26)
Amazon Bedrock shuts down ai21.jamba-1-5-mini-v1:0 on 2026-11-26.
Amazon Bedrock: anthropic.claude-3-7-sonnet-20250219-v1:0 shuts down in 17 days (2026-11-13)
Amazon Bedrock shuts down anthropic.claude-3-7-sonnet-20250219-v1:0 on 2026-11-13.
openai/codex shipped 60 releases in four weeks, faster than its usual pace
openai/codex published 60 releases in the last 28 days against 0 in the 56 days before.
vals-ai/valkyrie has shipped no release in four weeks after 16 in the eight before
vals-ai/valkyrie published 16 releases in the 56 days before the last four weeks and none since. A stall can mean a larger release in preparation or a project losing momentum.
cosmicstack-labs/mercury-agent shipped 15 releases in four weeks, faster than its usual pace
cosmicstack-labs/mercury-agent published 15 releases in the last 28 days against 6 in the 56 days before.
alexsergivan/kwant: 228 commits last week, manyx its usual pace
alexsergivan/kwant had 228 commits in the last complete week against a median of 0 before. A burst of work often precedes a release.
alexsergivan/kwant shipped 5 releases in four weeks, faster than its usual pace
alexsergivan/kwant published 5 releases in the last 28 days against 0 in the 56 days before.
nace-ai/kapitsa: 406 commits last week, manyx its usual pace
nace-ai/kapitsa had 406 commits in the last complete week against a median of 0 before. A burst of work often precedes a release.
nace-ai/kapitsa shipped 32 releases in four weeks, faster than its usual pace
nace-ai/kapitsa published 32 releases in the last 28 days against 0 in the 56 days before.
colinhacks/reylo shipped 8 releases in four weeks, faster than its usual pace
colinhacks/reylo published 8 releases in the last 28 days against 0 in the 56 days before.
ethanplusai/harvey: 43 commits last week, 43x its usual pace
ethanplusai/harvey had 43 commits in the last complete week against a median of 1 before. A burst of work often precedes a release.
cline/cline shipped 35 releases in four weeks, faster than its usual pace
cline/cline published 35 releases in the last 28 days against 25 in the 56 days before.
vercel/ai shipped 60 releases in four weeks, faster than its usual pace
vercel/ai published 60 releases in the last 28 days against 0 in the 56 days before.
langchain-ai/langgraph shipped 25 releases in four weeks, faster than its usual pace
langchain-ai/langgraph published 25 releases in the last 28 days against 9 in the 56 days before.
huggingface/transformers shipped 6 releases in four weeks, faster than its usual pace
huggingface/transformers published 6 releases in the last 28 days against 5 in the 56 days before.
morluto/rea shipped 15 releases in four weeks, faster than its usual pace
morluto/rea published 15 releases in the last 28 days against 1 in the 56 days before.
morluto/rea: 1,195 commits last week, 10.7x its usual pace
morluto/rea had 1,195 commits in the last complete week against a median of 112 before. A burst of work often precedes a release.
perplexityai/perplexity-py shipped 6 releases in four weeks, faster than its usual pace
perplexityai/perplexity-py published 6 releases in the last 28 days against 5 in the 56 days before.
google-gemini/gemini-cli shipped 32 releases in four weeks, faster than its usual pace
google-gemini/gemini-cli published 32 releases in the last 28 days against 28 in the 56 days before.
QwenLM/qwen-code shipped 46 releases in four weeks, faster than its usual pace
QwenLM/qwen-code published 46 releases in the last 28 days against 14 in the 56 days before.
Infini-AI-Lab/hermit: 60 commits last week, 4.8x its usual pace
Infini-AI-Lab/hermit had 60 commits in the last complete week against a median of 13 before. A burst of work often precedes a release.