GET /api/tags
Retrieve a list of all available models based on configured providers. Only models whose providers have valid API keys configured will be returned.Request
This endpoint accepts no parameters. Simply send a GET request:Response
Array of available model objects.
The model identifier used in API requests.
The model name (same as
name).ISO 8601 timestamp (generated at request time for compatibility).
Model size in bytes (placeholder value: 1000000000).
SHA-256 digest of the model name (for Ollama compatibility).
Example
Model Configuration
The available models are determined by:- Environment Variables: API keys for providers (OpenAI, Google, OpenRouter)
- models.json: Custom model configuration file (optional)
- Built-in Models: Default models when no
models.jsonexists
Built-in Models
If nomodels.json file exists, these built-in models are available:
- OpenAI:
gpt-4o-mini,gpt-4.1-mini,gpt-4.1-nano,gpt-4o - Google:
gemini-2.5-flash,gemini-2.5-flash-lite - OpenRouter:
deepseek-r1
Custom Model Configuration
Create amodels.json file in your working directory to customize available models:
models.json
Only models whose providers have valid API keys configured via environment variables will be returned. If a provider’s API key is missing, its models will be filtered out.
