Choose by capability
AI model capabilities
Find the AI models that support the API features your application needs. Explore function calling, web search, structured responses and more, then compare standard token prices, context windows and published intelligence scores.
Function calling
Connect model responses to your application’s functions and external tools.
Structured outputs
Compare models with schema-constrained responses for reliable application data.
JSON mode
Find models with a listed response-format option and compare their token costs.
Reasoning
Compare models with listed reasoning controls, token prices and published intelligence scores.
Built-in web search
Find AI models with a listed built-in search option for web-connected applications.
Log probabilities
Compare models that expose token-level log probabilities for evaluation and analysis.
Deterministic seed
Find models with a listed seed parameter for repeatability experiments.
Parallel tool calls
Discover models with a listed parallel-call control for multi-tool workflows.
Prompt caching
Compare models with listed implicit prompt caching and standard token prices.
Support is specific to the model and API
These directories use explicitly listed model features. Missing feature data is not treated as support, and support on one serving API does not guarantee availability on another. A low token price also does not establish capability quality. Check the exact version and API documentation, then evaluate your application’s own tasks.
Compare costs for your workload