Open Models in the dashboard to see enabled models and customer prices per million tokens. Search by name, model ID, provider or family; you can also filter by Open/Closed classification, provider, family and context size. Copy the model ID you want to use in your request.
Open/Closed describes the model classification, independently of its upstream provider. Supported open models may appear under SingleAPI branding; inference is provider-hosted. Existing model IDs and BYOK provider selection remain unchanged.
The model's context and output limits affect how much text you can send and receive. Prices can differ for input, cached input and output tokens. Check the current catalogue before relying on a model or price.
Your SingleAPI key controls which models your application can use. In API Keys, choose access to all enabled models or restrict a key to selected models. A restricted key with no selected models cannot call any model. See Manage API keys.
To list the models available to a particular key, use the OpenAI SDK after configuring client as in the quickstart:
for model in client.models.list().data:
print(model.id)Use one of those IDs in client.chat.completions.create(model=...). The model ID selects the provider; using the OpenAI SDK does not make every request an OpenAI provider request. For exact catalogue fields and model-list response shapes, see the API reference.