feat: add nvidia/z-ai/glm-5.3-flash

[guillaumegay13]

✨ What changed

• Adds z-ai/glm-5.3-flash (nvidia) to the catalog, params cloned from nvidia/gemma-4-31b-it.yaml.
• Hosted serving: the model is developed by z-ai (OpenRouter: z-ai/glm-5.3-flash); this entry documents nvidia's serving of it, probed against nvidia'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
temperature	min=0	✅
temperature	max=1	✅
temperature	range policed	❌	temperature=2 accepted — declared max 1 may be too low or unenforced
top_p	max=1	✅
top_p	range policed	✅	top_p=2 rejected
max_tokens	min=1	✅
max_tokens	max=16384	➖	too large for a non-streaming probe — upper bound from docs
max_tokens	range policed	❌	vendor reports max 1048576, YAML says 16384
frequency_penalty	min=-2	✅
frequency_penalty	max=2	✅
frequency_penalty	range policed	✅	frequency_penalty=4 rejected
presence_penalty	min=-2	✅
presence_penalty	max=2	✅
presence_penalty	range policed	✅	presence_penalty=4 rejected
seed	min=0	✅
seed	max=18446744073709552000	❌	Failed to deserialize the JSON body into the target type: invalid type: floating point 1.8446744073709552e+19, expected i64 at line 1 column 122
seed	range policed	✅	seed=36893488147419103000 rejected
stop	declared	➖	type string has no probe — value carried from sibling/docs, not verified

[vercel]

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

Project	Deployment	Actions	Updated
modelparams.dev	Ready	Preview	Sep 12, 2026 3:55am UTC