feat: add alibaba/qwen3.8-omni-flash

[guillaumegay13]

✨ What changed

• Adds qwen3.8-omni-flash (alibaba, api_key) to the catalog, params cloned from alibaba/qwen3.8-2.4t-a95b.yaml.

💭 Why

Alibaba Qwen announced it (new model) and the catalog doesn't list it yet.

📝 Notes

• Liveness ✅ only: the model answers on its real endpoint; unverified params are listed in the evidence table below.
• Source: https://qwen.ai/blog?id=qwen3.8-omni-flash
• Draft PR, auto-proposed by manifest-model-watch.

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=2	❌	'temperature' must be Float
temperature	range policed	✅	temperature=4 rejected
top_p	min=0	✅
top_p	max=1	✅
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 20, 2026 5:34am UTC