openai-native.ts

[Fastidius-au]

App Version

current codebase is using completions

API Provider

OpenAI

Model Used

GPT4.1s or any multimodal ongoing

Actual vs. Expected Behavior

https://github.com/RooVetGit/Roo-Code/blob/main/src/api/providers/openai-native.ts

From release emaila advice

While they’re available in both the Chat Completions and Responses APIs, for the richest experience, we recommend the Responses API. It supports reasoning summaries—the model’s thoughts stream while you wait for the final response—and enables smarter tool use by preserving the model’s prior reasoning between calls.

https://platform.openai.com/docs/api-reference/responses/object

seems like this is the new way and you need a way to deal better now that 4.1 is likely the go forward coders for a little while....can we just have a toggle for both options in the model select area. I expect architect is response and code is completions but feel free to discuss .

Detailed Steps to Reproduce

Relevant API Request Output

Additional Context

No response