GPT5 + model reasoning_effort support
Please explain the motivation behind the feature request. Reasoning effort is missing from Goose when configuring models and it is a very useful setting for especially for complex tasks such as coding
Describe the solution you'd like Ideally an ability to set in the chat at the bottom next to the model
Describe alternatives you've considered There are none
Additional context It is a very useful and critical settings for most AI backed workflows
- [x] I have verified this does not duplicate an existing feature request
thanks. this is all true. however, we don't support the responses API. we can revisit this after: https://github.com/block/goose/issues/5270 is delivered
Thank you, @DOsinga but this flag has always been part of chat completion API if I recall even before responses API had been released!
Please advise
Docs: https://platform.openai.com/docs/api-reference/chat/create#chat_create-reasoning_effort Working code sample:
curl https://api.openai.com/v1/chat/completions \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5",
"reasoning_effort": "high",
"messages": [
{ "role": "system", "content": "You are a highly skilled poem writer" },
{ "role": "user", "content": "Write a poem titled '\''What does the fox say?'\''" }
]
}'
I'd love to use Goose as a coding agent - it works better than all coding stuff on the market believe it or not BUT no having high reasoning is a bit of a bummer :-)
oops you are right