Models and pricing
Where to find every model, its inputs and its price.
Browse models
- opinference.com/models lists every model you can call, by modality.
- Each model has its own page at
opinference.com/models/{id}. It shows the model's price and, for a media model, every input it accepts with its type, default and allowed range. - opinference.com/prices puts every price on one page.
These pages are generated from the live catalog, so they are the reference for limits and prices. This site does not copy them.
List models from the API
GET /v1/models returns the models your key can use, each with its published price.
curl https://api.opinference.com/v1/models \
-H "Authorization: Bearer $OP_API_KEY"{
"object": "list",
"data": [
{
"id": "openai/gpt-oss-120b",
"object": "model",
"created": 0,
"owned_by": "openai",
"modality": "text",
"pricing": {
"price_version": "...",
"currency": "USD",
"rules": [
{ "unit": "input_tokens", "per": 1000000, "usd": "..." },
{ "unit": "output_tokens", "per": 1000000, "usd": "..." }
]
}
}
]
}GET /v1/models/{id} returns one model. A model id can contain a slash, as in
openai/gpt-oss-120b; send it as is.
| Field | Meaning |
|---|---|
id | The value to send as model |
owned_by | The model's author, not the endpoint that serves it |
modality | text, image, video, audio_tts (speech), audio_stt (transcription) or music |
pricing.rules | One rule per billed unit. usd is the price for per units (one unit when per is absent) |
pricing.price_version | The price list the rules come from |
inputs | Media models only: each accepted input with its type, default and range |
If a key is restricted to some models, the list holds only those. See Key restrictions.
How a request is priced
You pay the published price of the model you asked for, in the units its rules name: tokens for text, and seconds, images or characters for media. The cost of each request appears on your usage page. See Credits and billing.