feat: add openai/gpt-6-astra

[guillaumegay13]

✨ What changed

• Adds gpt-6-astra (openai) to the catalog, params cloned from openai/gpt-4.yaml.

💭 Why

The provider serves it but the catalog doesn't list it.

📝 Notes

• Liveness ✅ only: the model answers on its real endpoint; unverified params are listed in the evidence table below.
• Draft PR, auto-proposed by the modelparams agent.

Param verification evidence

param	check	ok	detail
(baseline)	accepted	✅	1-shot answers 200
max_completion_tokens	min=1	❌	Could not finish the message because max_tokens or model output limit was reached. Please try again with higher max_tokens.
max_completion_tokens	max=131072	➖	too large for a non-streaming probe — upper bound from docs
max_completion_tokens	range policed	❌	vendor reports max 128000, YAML says 131072
temperature	declared	➖	type number has no probe — value carried from sibling/docs, not verified
top_p	min=0	❌	Unsupported parameter: 'top_p' is not supported with this model.
top_p	max=1	✅
top_p	range policed	✅	top_p=2 rejected
reasoning_effort	=low	✅	accepted (no effect oracle for this value)
reasoning_effort	=medium	❌	ACCEPTED BUT NO EFFECT — response shows no sign the param was honored
reasoning_effort	=high	❌	ACCEPTED BUT NO EFFECT — response shows no sign the param was honored
reasoning_effort	enum policed	✅	vendor also accepts [xhigh] — YAML is missing them
response_format.type	=text	✅	accepted (no effect oracle for this value)
response_format.type	=json_schema	✅	accepted + effect confirmed in response
response_format.type	enum policed	✅	vendor also accepts [json_object] — YAML is missing them
tool_choice	=auto	❌	rejected: 400 Function tools with reasoning_effort are not supported for gpt-6-astra in /v1/chat/completions. To use function tools, use /v1/responses or set reasoning_effort
tool_choice	=none	❌	rejected: 400 Function tools with reasoning_effort are not supported for gpt-6-astra in /v1/chat/completions. To use function tools, use /v1/responses or set reasoning_effort
tool_choice	=required	❌	rejected: 400 Function tools with reasoning_effort are not supported for gpt-6-astra in /v1/chat/completions. To use function tools, use /v1/responses or set reasoning_effort
tool_choice	enum policed	✅	bogus value rejected

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
modelparams.dev	Ready	Preview	Sep 5, 2026 2:55am UTC