Models

SmartAPI exposes a catalog of models from multiple providers. You pass the model’s identifier in the model field of your request, and SmartAPI routes it to the appropriate upstream provider.

Browsing models#

The full, up-to-date catalog — including pricing — is available on the Models page. Each model lists its provider, type, supported endpoint protocols, and credit-based pricing.

Choosing an endpoint#

The model you pick determines which endpoint to call:

Model metadata#

Each model in the catalog carries the following attributes:

• Model name — the identifier you send in the model field (for example gpt-4.1, claude-haiku-4-6).
• Provider — the upstream provider, such as openai or anthropic.
• Type — the capability category: LLM, EMBEDDING, IMAGE_GENERATION, TEXT_TO_SPEECH, or VIDEO_GENERATION.
• Billing type — how usage is charged: USAGE (per token), PER_CALL, or MODEL (image/video billed in credits).
• Price — token-based models show credits per 1M tokens (input / output / cached input); image/video models show credit unit prices (per image / per second, plus resolution multipliers for video). All prices use the exchange rate $1 = 100 credits.
• Tags — capability markers; models tagged thinking support thinking/reasoning.

Example: list models programmatically#

The model catalog can also be fetched from the console API (no key required):

curl https://api.smartapi.cc/api/console/v1/model/list \
  -H "Content-Type: application/json" \
  -d '{}'

The response is wrapped in the standard console envelope ({ "code": 0, "message": "...", "data": ... }), where code equal to 0 indicates success.