- CBYellowstone2026-08-2124393
该提交已经包含在今天早上发布的149版本中。
github.com/openai/codex
Raise the GPT-5.6 maximum context window (#39102)
main←copyberry/codex-internal-to-codex-oss/9f530dee501222a4d280bb31bbcdc2dacbe35d7d已合并 11:09PM - 17 Aug 26 UTC

copyberry[bot]
+100
-39Raise the GPT-5.6 maximum context window
What changed
- Allow
gpt-5.6-sol…,gpt-5.6-terra, andgpt-5.6-lunacontext-window overrides up to 872,000 tokens. - Build the corresponding Amazon Bedrock entries from each bundled model's metadata while preserving Bedrock-specific identifiers, availability, tool, reasoning, and service-tier settings.
Testing
- Cover long-context override clamping for bundled and Amazon Bedrock GPT-5.6 models.
- Verify the normalized Amazon Bedrock catalog metadata and context limits.
主楼 - Allow
- Kevin-linuxdo2026-08-216133
默认800k了?超过257k还会双倍扣费么
- alizoed2026-08-2113055
不知道用 ccs+cpa 的什么时候可以默认这个上下文
- FFFFFFFFFir2026-08-2118888
直接拉这么大性能劣化情况怎么样呢,5.5不提,5.4可是惨不忍睹
- christ2026-08-2139571
我比较关心的是800k,性能真不要了?
- OneDingZ2026-08-2123833
超过两三百k的性能就下滑得很厉害了
- bige01232026-08-2115672
我设置在400K, 超过400K我不是很相信他的长上下文的性能
- blacksein2026-08-219329
- CBYellowstone2026-08-2124393
- 91kevinshi2026-08-2123383
我觉得可以暂时不管计费问题,主要是372000太难用了额
- cut2026-08-2138778
接中转站也是嘛?刚才更新了149但是还是258k,需要怎么设置佬们
- GloryDuck2026-08-218299
- W1nge2026-08-214840
