[Souravrajvi0]

Summary

Fixes #3827.

The Chat Completions path popped reasoning_effort from model parameters but only re-added it when enable_thinking was explicitly true. In the default "Both Modes" configuration (thinking toggle off), the standard OpenAI parameter was silently dropped from outgoing requests.

This change forwards the top-level reasoning_effort whenever thinking is not forced off (enable_thinking_value is not False), matching the Responses API path and the issue's suggested fix. The non-standard chat_template_kwargs mirror remains gated on extended compatibility mode and an explicit thinking toggle.

Bump plugin version to 0.0.66.

Release Notes

• Fix Chat Completions requests dropping the standard reasoning_effort parameter when the thinking toggle is off in "Both Modes".
• Forward reasoning_effort whenever thinking is not explicitly forced off, matching the Responses API behavior.
• Keep the extended chat_template_kwargs mirror gated on extended compatibility mode and an explicit thinking toggle.

Change Type

• Documentation / non-plugin change
• Non-LLM plugin (tools, extensions, datasource, etc.)
• LLM plugin

Test plan

• Added test_reasoning_effort_forwarding.py covering:
  • default "Both Modes" with thinking toggle off forwards reasoning_effort
  • thinking toggle on still forwards reasoning_effort
  • thinking forced off (not_supported) still drops reasoning_effort
  • thinking forced on (only_thinking_supported) forwards reasoning_effort
  • extended mode mirrors to chat_template_kwargs only when thinking is explicitly enabled
• uv run pytest tests/test_reasoning_effort_forwarding.py -v passes (6/6)
• After merging main, local tests on the updated branch pass

[Mairuis]

Please fix CI errors. @Souravrajvi0

[Souravrajvi0]

@Mairuis CI was failing at Check Release Notes. I added a ## Release Notes section, merged main, bumped the plugin version to 0.0.66 (since 0.0.65 was already on main), and pushed the update. Please rerun the checks if they do not start automatically.

[crazywoola]

Thank you for your contribution, @Souravrajvi0! This PR currently has merge conflicts with main, so we are closing it for now. Please update your branch against the latest main and resolve the conflicts — note that the consolidated dependency refresh (#4106–#4111) has bumped plugin manifest.yaml versions on main, so please bump models/openai_api_compatible/manifest.yaml to the next unused version when you rebase. You are welcome to open a new PR with the updated changes and link back to this one. We appreciate your time and effort!

If these changes are still wanted, they can be folded into #3869 when you update it.