feat: add alibaba/kimi-k3

[guillaumegay13]

✨ What changed

• Adds kimi-k3 (alibaba) to the catalog, params cloned from alibaba/kimi-k2.7-code.yaml.
• Hosted serving: the model is developed by moonshotai (OpenRouter: moonshotai/kimi-k3); this entry documents alibaba's serving of it, probed against alibaba's endpoint.

💭 Why

The provider serves it but the catalog doesn't list it. Param surfaces differ per serving provider, so each host gets its own probed entry (models.dev-style).

📝 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	✅
max_completion_tokens	range policed	✅	no declared max in YAML; vendor accepts 100M (caps silently) — consistent
temperature	min=0	✅
temperature	max=1.9	❌	<400> InternalError.Algo.InvalidParameter: Parameter 'temperature'=1.9 is not supported for kimi-k3 model.
temperature	range policed	✅	temperature=3.8 rejected
top_p	min=0	❌	<400> InternalError.Algo.InvalidParameter: Parameter 'top_p'=0.0001 is not supported for kimi-k3 model.
top_p	max=1	❌	<400> InternalError.Algo.InvalidParameter: Parameter 'top_p'=1.0 is not supported for kimi-k3 model.
top_p	range policed	✅	top_p=2 rejected
extra_body.top_k	min=0	✅
extra_body.enable_thinking	=true	✅	accepted (no effect oracle)
extra_body.thinking_budget	min=1	✅

[vercel]

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

Project	Deployment	Actions	Updated
modelparams.dev	Ready	Preview	Sep 2, 2026 5:04pm UTC