Skip to main content

​
What are Providers?

Providers are the AI model services that power your agents. Havoc supports a wide range of providers, from major cloud services to local models.

​
Supported Providers

ProviderModelsNotes
AnthropicClaude Opus 4.6, Claude Sonnet 4.5Recommended for most use cases
OpenAIGPT-5.2, GPT-5-miniBroad capability
GoogleGemini 3 Pro, Gemini 3 FlashFast and cost-effective
OpenRouter100+ modelsMulti-provider gateway
xAIGrokAlternative provider
GroqLLaMA, MixtralUltra-fast inference
MistralMistral Large, Medium, SmallEuropean provider
OllamaAny local modelSelf-hosted, no API key needed
Together AIOpen-source modelsCost-effective
Amazon BedrockMultiple providersAWS-integrated

​
Built-in Aliases

For convenience, Havoc provides shorthand aliases:
AliasResolves to
opusanthropic/claude-opus-4-6
sonnetanthropic/claude-sonnet-4-5
gptopenai/gpt-5.2
gpt-miniopenai/gpt-5-mini
geminigoogle/gemini-3-pro-preview
gemini-flashgoogle/gemini-3-flash-preview

​
Model Configuration

Each agent can configure a primary model with fallbacks: 
{
  "model": {
    "primary": "anthropic/claude-sonnet-4-5",
    "fallbacks": ["openai/gpt-5.2"]
  },
  "models": {
    "anthropic/claude-sonnet-4-5": { "alias": "Sonnet" },
    "openai/gpt-5.2": { "alias": "GPT" }
  }
}
If the primary model is unavailable or rate-limited, the agent automatically falls back to the next model in the list.

​
Custom Providers

You can connect any OpenAI-compatible endpoint as a custom provider: 
{
  "providers": {
    "my-proxy": {
      "baseUrl": "http://localhost:4000/v1",
      "apiKey": "your-key",
      "api": "openai-completions",
      "models": [
        {
          "id": "custom-model",
          "name": "My Custom Model",
          "contextWindow": 128000,
          "maxTokens": 32000
        }
      ]
    }
  }
}
Supported API types: openai-completions, openai-responses, anthropic-messages, google-generative-ai.

​
Thinking & Reasoning

Agents support configurable reasoning levels:
LevelDescription
offNo reasoning
minimalBrief internal reasoning
low (default)Standard reasoning
mediumExtended reasoning
highDeep reasoning
xhighMaximum reasoning (GPT-5.2/Codex only)

Guide: Configure a Provider

Step-by-step provider setup with API key management.