← Back
Anthropic
Anthropic adds model capability fields to Claude API, enabling developers to discover model features programmatically

Models API Enhanced with Capability Discovery

Anthropic has updated the Claude Platform's Models API to provide structured metadata about model capabilities directly through the API.

What Changed:

  • GET /v1/models and GET /v1/models/{model_id} endpoints now return three new fields:
    • max_input_tokens - maximum input context size for the model
    • max_tokens - maximum output tokens the model can generate
    • capabilities object - structured data about model features

Why This Matters:

Previously, developers had to consult documentation to understand model capabilities and limits. This change enables programmatic discovery, allowing applications to dynamically adapt to model features and select the most appropriate model for a task based on real-time capability data.

How to Use:

Query the Models API as you normally would—no authentication or configuration changes required. The response will include the new fields alongside existing model metadata. This is particularly useful for applications that need to verify capability support before making requests or for building abstraction layers that adapt to available features.