---
type: "post"
slug: "model-routing-and-pricing-update-glm-deepseek-grok-qwen"
title: "Model routing and pricing update: GLM, DeepSeek, Grok, Qwen"
description: "Model routing and pricing update: GLM, DeepSeek, Grok, Qwen ties several source-backed updates into one New Runtime pattern for agent and AI infrastructure work."
retrieval_nugget: "Model routing and pricing update: GLM, DeepSeek, Grok, Qwen ties several source-backed updates into one New Runtime pattern for agent and AI infrastructure work. Four narrow updates show model selection becoming a gateway and harness economics problem: GLM appears in Vercel AI Gateway, OpenCode reacts to DeepSeek pricing, Grok 4.6 reaches more coding surfaces,"
published_at: "2026-08-20"
updated_at: "2026-08-20"
record_date: "2026-08-20"
date_kind: "published_at"
topics: ["ai-news","agents","models","developer-tools"]
source_cluster_ref: null
source_synthesis_id: "e77e79d3-b7de-4687-9e37-0eb8b3c51293"
source_synthesis_op_ref: "model-routing-price"
assigned_agent: "hermes"
basket_id: "bb845773-3bce-437b-aae5-d91c181b7d21"
basket_revision: 1
body_word_count: 156
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:model-routing-and-pricing-update-glm-deepseek-grok-qwen"
status: "published"
source_urls: ["https://x.com/vercel_dev/status/2089123749572616645","https://x.com/vercel_dev/status/2089123749572616645/photo/1","https://x.com/opencode/status/2089428008268402786","https://x.com/opencode/status/2089030951653240880","https://x.ai/news/grok-4-6-github-copilot","https://x.com/augmentcode/status/2088366283020759207","https://www.augmentcode.com","https://x.com/grok/status/2088319896597975501","https://huggingface.co/Qwen/Qwen3.8-27B"]
visuals: []
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:26fc7a4e56939ec84485aaab88aa93e7e61ab6ac002685c496bb0be7be9481a3","reviewed_at":"2026-08-20T19:10:59.652Z","source_evidence_count":1,"verified_claim_count":2}
routes: {"html":"https://newruntime.com/posts/model-routing-and-pricing-update-glm-deepseek-grok-qwen/","markdown":"https://newruntime.com/posts/model-routing-and-pricing-update-glm-deepseek-grok-qwen.md","json":"https://newruntime.com/posts/model-routing-and-pricing-update-glm-deepseek-grok-qwen.json"}
---

# Model routing and pricing update: GLM, DeepSeek, Grok, Qwen

## Retrieval answer

Model routing and pricing update: GLM, DeepSeek, Grok, Qwen ties several source-backed updates into one New Runtime pattern for agent and AI infrastructure work. Four narrow updates show model selection becoming a gateway and harness economics problem: GLM appears in Vercel AI Gateway, OpenCode reacts to DeepSeek pricing, Grok 4.6 reaches more coding surfaces,

Model routing and pricing update: GLM, DeepSeek, Grok, Qwen is a synthesis post, not a separate claim that every linked item has the same weight. The useful signal is the shared direction: Four narrow updates show model selection becoming a gateway and harness economics problem: GLM appears in Vercel AI Gateway, OpenCode reacts to DeepSeek pricing, Grok 4.6 reaches more coding surfaces, and Qwen3.8-27B appears on Hugging Face.

The source evidence for this site row is one accountable public source from the synthesis set, while the remaining links stay attached so a reader or agent can inspect the wider pattern. That keeps the post useful without pretending that a roundup is a single-source release note.

The operational takeaway is simple: the newsroom should track these smaller updates as part of a capability layer. Each individual item may be narrow, but together they change how teams route models, package agent skills, verify output, or govern infrastructure boundaries.