Llm providers

For each model provider, you can configure parameters either through environment variables or in your fast-agent.yaml file.

Be sure to run fast-agent check to troubleshoot API Key issues:

Key Check

Common Configuration Format

In your fast-agent.yaml:

<provider>:
  api_key: "your_api_key" # Override with API_KEY env var
  base_url: "https://api.example.com" # Base URL for API calls
  default_headers: # Optional - custom headers for all API requests
    X-Custom-Header: "value"

The default_headers option is available for OpenAI-compatible providers (including Azure).

Provider quick reference

Provider	Config key	API key environment variable	Notes
OpenAI Responses	`responses`	`OPENAI_API_KEY`	Native Responses API, reasoning, web search, connectors, remote MCP
OpenAI Chat Completions	`openai`	`OPENAI_API_KEY`	OpenAI-compatible Chat Completions models
Codex Responses	`codexresponses`	`fast-agent auth codexplan` or `CODEX_API_KEY`	Codex subscription-backed Responses models; no provider-managed MCP/connectors
Anthropic	`anthropic`	`ANTHROPIC_API_KEY`	Claude Messages API, prompt caching, web tools
Google	`google`	`GOOGLE_API_KEY`	Native Gemini API
Azure OpenAI	`azure`	`AZURE_OPENAI_API_KEY`	Azure deployments and optional DefaultAzureCredential
Bedrock	`bedrock`	AWS credentials	Amazon Bedrock models
Groq	`groq`	`GROQ_API_KEY`	Additional provider; OpenAI-compatible hosted inference
xAI / Grok	`xai`	`XAI_API_KEY`	Grok models, reasoning, web search, and X Search
DeepSeek	`deepseek`	`DEEPSEEK_API_KEY`	Additional provider; OpenAI-format API
Aliyun	`aliyun`	`ALIYUN_API_KEY`	Additional provider; DashScope compatible-mode endpoint
OpenRouter	`openrouter`	`OPENROUTER_API_KEY`	Additional provider; routed upstream models
Hugging Face	`hf` or `huggingface`	`HF_TOKEN`	Inference Providers router and HF MCP auth
Open Responses	`openresponses`	`OPENRESPONSES_API_KEY`	Additional provider; interoperable Open Responses endpoints
Generic	`generic`	`GENERIC_API_KEY`	Additional provider; local/self-hosted OpenAI-compatible endpoints
TensorZero	`tensorzero`	None	Additional provider; local TensorZero Gateway functions

See Additional Providers for the long-tail reference with config keys, API key names, default endpoints, model string examples, and provider-specific notes.

Detailed provider guides

OpenAI for Responses, Chat Completions, Codex Responses, web tools, connectors, and transport options.
Anthropic for Claude, prompt caching, reasoning, structured outputs, and Anthropic web tools.
Google for native Gemini configuration and aliases.
Azure OpenAI for Azure deployments, authentication modes, and regional availability.
AWS Bedrock for Bedrock model IDs, AWS authentication, and capability caveats.
xAI / Grok for Grok models, reasoning, web search, and X Search.
Hugging Face for Inference Providers routing, curated aliases, and HF MCP authentication.
Additional Providers for Groq, DeepSeek, Aliyun, OpenRouter, Open Responses, TensorZero, and generic OpenAI-compatible endpoints.