fix: normalize minimal effort for documented OpenAI models - #59
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
reasoning_effort="minimal"and small positive thinking budgets produced unsupportedminimalrequests for GPT-5.4 Mini and GPT-5.5 on both OpenAI API-key and OAuth paths. Their documented least positive effort islow.A shared compatibility helper now selects
lowfor the verified GPT-5.1, GPT-5.2, GPT-5.4, GPT-5.4 Mini, GPT-5.4 Nano and GPT-5.5 aliases and their documented snapshots. Both adapters use it. The existing Codex rule and Azure override remain intact; legacy GPT-5 keeps minimal, and unset controls still omit effort. Exact model matching avoids making assumptions about pro variants or future models. No new effort values or model catalog entries, including no expansion of the OAuth model catalog.Validation: 274 targeted tests passed with four xdist workers. New regressions exercise real BaseAPIModel sync/async calls for both providers, capture final transport payloads, and cover documented aliases/snapshots. Nano invocation coverage uses the API-key path. Against the unchanged merged baseline, 36 new cases failed on minimal instead of low, with 82 passing controls; the additional nano cases separately failed before the two-ID addition (8 failures, 10 controls passing). Related OAuth forwarding, cache-key, Azure, schema, token-refresh and streaming tests pass. No live model calls or credential reads.
Model contracts: GPT-5.1, GPT-5.2, GPT-5.4, GPT-5.4 Mini, GPT-5.4 Nano, GPT-5.5.