GET /v1/models

List available models

Endpoint

GET /v1/models

Retrieve a list of all models available across your configured accounts.

Request

No parameters required. This is a simple GET request.

curl http://localhost:8080/v1/models

Response

object · string

Always "list".

data · array

Array of model objects. Each model has:

• id (string): Model identifier (e.g., "claude-sonnet-4-5-thinking")
• object (string): Always "model"
• created (number): Timestamp when model was added
• owned_by (string): Model provider ("anthropic" or "google")

Example Response

{
  "object": "list",
  "data": [
    {
      "id": "claude-opus-4-6-thinking",
      "object": "model",
      "created": 1704067200,
      "owned_by": "anthropic"
    },
    {
      "id": "claude-sonnet-4-5-thinking",
      "object": "model",
      "created": 1704067200,
      "owned_by": "anthropic"
    },
    {
      "id": "claude-sonnet-4-5",
      "object": "model",
      "created": 1704067200,
      "owned_by": "anthropic"
    },
    {
      "id": "gemini-3.1-pro-high",
      "object": "model",
      "created": 1704067200,
      "owned_by": "google"
    },
    {
      "id": "gemini-3.1-pro-low",
      "object": "model",
      "created": 1704067200,
      "owned_by": "google"
    },
    {
      "id": "gemini-3-flash",
      "object": "model",
      "created": 1704067200,
      "owned_by": "google"
    }
  ]
}

Model Details

Claude Models

claude-opus-4-6-thinking

Provider: Anthropic
Capabilities: Extended reasoning, highest intelligence
Best for: Complex problem-solving, deep analysis
Thinking: Supported
Max tokens: 4096 (configurable via max_tokens)

claude-sonnet-4-5-thinking

Provider: Anthropic
Capabilities: Balanced reasoning and speed
Best for: General-purpose tasks with reasoning
Thinking: Supported
Max tokens: 4096 (configurable via max_tokens)

claude-sonnet-4-5

Provider: Anthropic
Capabilities: Fast, production-ready
Best for: Quick tasks, production workloads
Thinking: Not supported
Max tokens: 4096 (configurable via max_tokens)

Gemini Models

gemini-3.1-pro-high

Provider: Google
Capabilities: Highest quality Gemini
Best for: Complex tasks requiring high accuracy
Thinking: Supported (Gemini 3+)
Max tokens: 16384 (Gemini limit)

gemini-3.1-pro-low

Provider: Google
Capabilities: Cost-optimized Gemini
Best for: Budget-conscious tasks
Thinking: Supported (Gemini 3+)
Max tokens: 16384 (Gemini limit)

gemini-3-flash

Provider: Google
Capabilities: Fastest Gemini model
Best for: Low-latency tasks, real-time responses
Thinking: Supported (Gemini 3+)
Max tokens: 16384 (Gemini limit)

Model Availability

Model availability depends on your configured accounts:

• Free tier: Access to all Claude and Gemini models
• Pro tier: Higher quotas and rate limits
• Ultra tier: Highest quotas and priority access

The /v1/models endpoint queries the first available account to fetch the model list. If no accounts are available, it returns a 503 error.

Model Metadata

The proxy dynamically discovers models from the Antigravity Cloud Code API using the fetchAvailableModels endpoint. This ensures the model list is always up-to-date with the latest models available to your accounts.

Model Mapping

You can configure model aliases in config.json to map model names:

{
  "modelMapping": {
    "claude-3-5-sonnet-20241022": {
      "mapping": "claude-sonnet-4-5-thinking",
      "hidden": false
    }
  }
}

This allows you to use legacy model names in your client while the proxy routes to the correct underlying model.

Error Response

503 Service Unavailable

Returned when no accounts are available:

{
  "type": "error",
  "error": {
    "type": "api_error",
    "message": "No accounts available"
  }
}

500 Internal Server Error

Returned when the model list cannot be fetched:

{
  "type": "error",
  "error": {
    "type": "api_error",
    "message": "Failed to fetch models: <error details>"
  }
}