gpt-6.1 sol context limit is smaller in v2 #54141

Activity

1. github-actions

   assigned

   neriousy

   on Oct 9, 2026
2. opencode-agent commented on Oct 9, 2026

   

   opencode-agentBot

   on Oct 9, 2026 – with OpenCode Agent

   Contributor

   More actions

   Thanks for the report! I think this is a duplicate of #47646.

   When OpenAI is connected through ChatGPT OAuth, v2 replaces the model's catalog limit (about 1.05M context) with a fixed 400k context / 272k input, which matches the Codex CLI's budget. That applies to gpt-6.1-sol and the other gpt-6 models.

   Possible duplicates and related items, ordered by likelihood:

   • openai: ChatGPT OAuth reports 400k context for long-context models #47646: ChatGPT OAuth reports 400k context for long-context models on v2. It names gpt-6.1-sol, and another user notes v1 had the full 1M window.
   • fix(openai): OAuth transform treats Codex product budget as endpoint context limit #44821: same root cause, with measurements showing the endpoint accepts far more input. Its comments include a workaround: override limit per model under provider.openai.models in your opencode.json. Someone confirmed it works for gpt-6.1-sol on v2.
   • fix(core): preserve OpenAI OAuth model context limits #53080: an open PR that proposes removing the fixed 400k limit.

   If you're using an OpenAI API key rather than ChatGPT OAuth, or your problem is different in some other way, please reply here and explain. Otherwise the bot will close this issue as a duplicate in 1 day.

   👍React with 👍1Reacted by Alexander Petrov
3. opencode-agent

   added

   pending close

   triaging

   on Oct 9, 2026
4. zowers

   closed this as a duplicateof openai: ChatGPT OAuth reports 400k context for long-context models #47646 on Oct 9, 2026
5. opencode-agent commented on Oct 10, 2026

   

   opencode-agentBot

   on Oct 10, 2026 – with OpenCode Agent

   Contributor

   More actions

   Closing this as a duplicate of #47646, since there's been no reply saying the problem is different. In v2, ChatGPT OAuth connections cap gpt-6.1-sol and the other gpt-6 models at 400k context / 272k input instead of their catalog limit.

   You can follow progress in #47646 and the open fix in PR #53080. Until then, #44821 describes a workaround: override limit per model in your opencode.json.

   If you're using an OpenAI API key rather than ChatGPT OAuth, or your case is different in some other way, please comment here and we can reopen. Thanks for the report!