Skip to main content
The Ollama API Proxy supports multiple AI providers and models. You can customize the available models by editing the models.json file in your project root.

​
Model configuration

Models are defined in models.json with the following structure: 
{
  "model-name": {
    "provider": "openai",
    "model": "actual-model-id"
  }
}
  • model-name: The name clients use to request this model (e.g., gpt-4o-mini)
  • provider: The AI provider (openai, google, or openrouter)
  • model: The actual model identifier used by the provider
Models are only available if their corresponding provider API key is configured in your .env file.

​
Default models

The following models are available by default:

​
OpenAI models

Requires OPENAI_API_KEY in your environment.
Provider: OpenAI
Model ID: gpt-4o-mini
Description: Fast, cost-effective model suitable for most tasks
{
  "provider": "openai",
  "model": "gpt-4o-mini"
}
Provider: OpenAI
Model ID: gpt-4.1-mini
Description: Latest GPT-4.1 mini variant
{
  "provider": "openai",
  "model": "gpt-4.1-mini"
}
Provider: OpenAI
Model ID: gpt-4.1-nano
Description: Ultra-fast, lightweight model
{
  "provider": "openai",
  "model": "gpt-4.1-nano"
}
Provider: OpenAI
Model ID: gpt-4o
Description: Full GPT-4o model with vision support
{
  "provider": "openai",
  "model": "gpt-4o"
}
Provider: OpenAI
Model ID: gpt-5-nano
Description: Next-generation lightweight model
{
  "provider": "openai",
  "model": "gpt-5-nano"
}

​
Google models

Requires GEMINI_API_KEY in your environment.
Provider: Google
Model ID: gemini-2.5-flash
Description: Fast Gemini model with multimodal capabilities
{
  "provider": "google",
  "model": "gemini-2.5-flash"
}
Provider: Google
Model ID: gemini-2.5-flash-lite
Description: Lightweight version of Gemini Flash
{
  "provider": "google",
  "model": "gemini-2.5-flash-lite"
}

​
OpenRouter models

Requires OPENROUTER_API_KEY in your environment.
Provider: OpenRouter
Model ID: deepseek/deepseek-r1-0528:free
Description: Free DeepSeek reasoning model
{
  "provider": "openrouter",
  "model": "deepseek/deepseek-r1-0528:free"
}
Provider: OpenRouter
Model ID: moonshotai/kimi-k2:free
Description: Free Kimi K2 model from Moonshot AI
{
  "provider": "openrouter",
  "model": "moonshotai/kimi-k2:free"
}

​
Adding custom models

You can add any model supported by your providers:
1

Edit models.json

Open the models.json file in your project root.
2

Add your model configuration

Add a new entry with your desired model name:
{
  "my-custom-model": {
    "provider": "openai",
    "model": "gpt-4-turbo"
  }
}
3

Restart the proxy

Restart the Ollama API Proxy to load the new configuration:
npm start
4

Verify the model is available

Check the startup logs to confirm your model is loaded:
📋 Available models: gpt-4o-mini, ..., my-custom-model
For OpenRouter models, you can browse the full catalog at openrouter.ai/models and use the model ID in your configuration.

​
Provider setup

Each provider requires an API key configured in your .env file: 
# OpenAI models
OPENAI_API_KEY=sk-your-openai-key

# Google Gemini models
GEMINI_API_KEY=your-gemini-key

# OpenRouter models (access to multiple providers)
OPENROUTER_API_KEY=sk-or-your-openrouter-key
OPENROUTER_API_URL=https://openrouter.ai/api/v1  # Optional, uses default if not set
The proxy will exit with an error if no API keys are configured. You must set at least one provider API key.

​
Checking available models

You can query the proxy to see which models are currently available: 
curl http://localhost:11434/api/tags
Response: 
{
  "models": [
    {
      "name": "gpt-4o-mini",
      "model": "gpt-4o-mini",
      "modified_at": "2026-03-11T10:30:00.000Z",
      "size": 1000000000,
      "digest": "sha256:gpt4omini"
    }
    // ... more models
  ]
}

​
Next steps

Vision Support

Learn how to use vision-capable models

JetBrains AI Assistant

Configure JetBrains IDE to use the proxy

Build docs developers (and LLMs) love

Get started for freeTalk to us