Models catalog
What the gateway serves, and what it costs.
The catalog below is what this deployment serves. Your key reaches exactly these ids, and GET /v1/models returns the same list at runtime.
| Model id | Context | Capabilities | Input / 1M | Cached | Output / 1M |
|---|---|---|---|---|---|
unrefused-glm-5.3-flash · default | 1M | chat, completions, tools | $3.00 | $0.30 | $5.00 |
embed-base | 8K | embeddings | $0.02 | — | — |
Placeholder rates while capacity is finalised. Published prices live on the pricing page.
Switching models
Model ids are stable strings. Change model and nothing else in your request changes: the shape, tool calling, streaming and the policy decisions behave identically across the catalog.
client.chat.completions.create(
model="unrefused-glm-5.3-flash",
messages=msgs,
)Asking at runtime
Prefer not to hard-code? Call GET /v1/models and read context_window and capabilities before dispatching. Unknown ids return 404 not_found with param: "model".
Spot an error or a gap? Tell us — and quote x-request-id if you have one.