Switch to LiteLLM Responses API; default gpt-5 and --fast gpt-5-nano with high reasoning#3

Open

devin-ai-integration[bot] wants to merge 3 commits intomainfrom

devin/1757769949-switch-to-litellm-responses-gpt5

Commits on Sep 13, 2025

feat(models): default gpt-5; --fast uses gpt-5-nano; switch to litellm.responses with high reasoning

devin-ai-integration[bot]
and
shouryamaanjain
committed
docs: update user-facing prompts to reference GPT-5 instead of GPT-4o

devin-ai-integration[bot]
and
shouryamaanjain
committed
responses: enable streaming and wrap events into OpenAI-style chunks; update Azure prompt to GPT-5

devin-ai-integration[bot]
and
shouryamaanjain
committed