Providers

Schema Reference: For the most up-to-date and complete provider configuration schema, see the Provider Schema Documentation. The schema contains all available properties, validation rules, and provider-specific requirements.

Supported Providers

1
OpenAI
GPT-4, GPT-3.5, and other OpenAI models
2
Anthropic
Claude 3 family models for advanced reasoning
3
Google
Gemini Pro and other Google AI models
4
Groq
Fast inference platform with OpenAI-compatible API
5
Ollama
Local model hosting for self-hosted models
6
DeepSeek
DeepSeek AI models with OpenAI-compatible API
7
xAI
Grok models with OpenAI-compatible API

Basic Configuration

Configure providers in the models section of your compozy.yaml:

models:
  - provider: openai
    model: gpt-4
    api_key: "{{ .env.OPENAI_API_KEY }}"

  - provider: anthropic
    model: claude-3-5-sonnet-20241022
    api_key: "{{ .env.ANTHROPIC_API_KEY }}"

  - provider: google
    model: gemini-pro
    api_key: "{{ .env.GOOGLE_API_KEY }}"

Configuration Reference

Core Properties

All providers support these core configuration properties. For the complete schema definition with all available fields and validation rules, see the Provider Schema Documentation.

Property	Type	Required	Description
`provider`	string	✅	Provider name (see supported providers above)
`model`	string	✅	Model identifier - any valid model string for the provider
`api_key`	string	⚠️	API key for authentication (required for cloud providers)
`api_url`	string	❌	Custom API endpoint URL
`organization`	string	❌	Organization ID (OpenAI only)
`params`	object	❌	Generation parameters (temperature, max_tokens, etc.)

Parameter Support

Parameters are specified in the params object. The available parameters vary by provider - consult the Provider Schema Documentation for provider-specific parameter schemas:

models:
  - provider: openai
    model: gpt-4
    api_key: "{{ .env.OPENAI_API_KEY }}"
    params:
      temperature: 0.7
      max_tokens: 4000
      top_p: 0.9
      seed: 42

Parameter Compatibility: Not all parameters are supported by all providers. Unsupported parameters are either ignored or cause configuration errors depending on the provider. Check the Provider Schema Documentation for detailed parameter compatibility.

Provider-Specific Configuration

OpenAI Configuration

models:
  - provider: openai
    model: gpt-4
    api_key: "{{ .env.OPENAI_API_KEY }}"
    organization: "org-123"                # Optional
    params:
      temperature: 0.7
      max_tokens: 4000
      top_p: 0.9
      seed: 42

Model Support: Any valid OpenAI model string is supported. Popular models include:

gpt-4 - Most capable model
gpt-4-turbo - Faster GPT-4 variant
gpt-3.5-turbo - Faster, cheaper option
gpt-4o - Multimodal capabilities
And any other model available in your OpenAI account

Error Conditions:

✅ All configuration options supported
✅ Organization parameter supported
✅ Custom API URL supported

Environment Variables

Store API keys in environment variables, never in configuration files.

# .env
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...
GOOGLE_API_KEY=AIza...
GROQ_API_KEY=gsk_...
DEEPSEEK_API_KEY=sk-...
XAI_API_KEY=xai-...

Security Best Practices

API Key Management
Use environment variables for API keys and rotate them regularly
Network Security
Use HTTPS endpoints and verify SSL certificates
Access Control
Implement proper authentication and authorization
Input Validation
Validate and sanitize all inputs before sending to providers
Error Handling
Handle provider errors gracefully and avoid exposing sensitive information

Multiple Provider Configuration

You can configure multiple providers and models:

models:
  # Primary provider
  - provider: openai
    model: gpt-4
    api_key: "{{ .env.OPENAI_API_KEY }}"
    params:
      temperature: 0.7

  # Alternative provider
  - provider: anthropic
    model: claude-3-5-sonnet-20241022
    api_key: "{{ .env.ANTHROPIC_API_KEY }}"
    params:
      temperature: 0.5

  # Local provider for development
  - provider: ollama
    model: llama2:7b
    api_url: "http://localhost:11434"
    params:
      temperature: 0.8

Global Configuration

Configure system-wide settings and application defaults

Runtime Configuration

Set up execution environments and tool runtime settings

Agent Configuration

Configure AI agents and their behavior

Workflow Configuration

Define workflow orchestration and task management

On this page