Available Models
Discover current models, providers, and protocols from the public API
OpenModel provides access to models from multiple AI providers through a unified API. This guide helps you inspect current providers and protocols and find a suitable model. Models, providers, and protocols come from the public API rather than a business list maintained in these docs.
Client deals
Some models advertise benefits for supported clients such as Codex or Claude Code. Model Pricing shows currently published offers, while the consumption log shows the price actually applied to a request. Learn more in Client Benefits.
Current providers and protocols
Protocols
OpenModel currently provides the following API protocols. Use the dynamic data above or Model Pricing as the source for the protocols supported by each model.
Messages Protocol (/v1/messages)
Use the Anthropic SDK with this protocol. Models that support it share the same endpoint and request format; switching models usually requires only changing the model parameter.
Responses Protocol (/v1/responses)
Use the OpenAI SDK with this protocol.
Gemini Protocol (/v1beta/models/{model}:generateContent)
Use the Google GenAI SDK with this protocol.
Discovering Models
In the Console
Model Pricing reads the public API directly and lists current models, prices, providers, protocols, and published client discounts. You can also use the Models page in the Console after signing in.
Via the Web API
Query the public models endpoint (no authentication required):
curl "https://api.openmodel.ai/web/v1/models"SDK Model List Endpoint
For the OpenAI SDK and tools that can query GET /v1/models independently of a generation endpoint:
curl https://api.openmodel.ai/v1/models \
-H "Authorization: Bearer $OPENMODEL_API_KEY"This endpoint only lists enabled models. Its existence does not imply support for Chat Completions or arbitrary OpenAI generation formats; use each model's published supported_protocols for requests.
Choosing a Protocol
Find the model in Model Pricing and inspect its current supported_protocols. A model can support more than one protocol. Choose the protocol that matches your SDK and request format instead of inferring it from the provider name.
Model Naming
Models are referenced by their key (model name string) as listed in the Console and the models API. Use this key as the model parameter in your API requests.
Model names generally follow the upstream provider's naming convention. Copy the current model key from the public API, Model Pricing, or the Console rather than inferring it from documentation examples.