Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.wednesdayai.dev/llms.txt

Use this file to discover all available pages before exploring further.

​
LiteLLM / Bifrost

Both LiteLLM and Bifrost are self-hosted LLM proxy servers that sit between WednesdayAI and AI providers. They provide centralized model management, budget limits, usage tracking, and team access control.

​
When to use a proxy

  • You run WednesdayAI for a team and want centralized API key management
  • You need per-user or per-team cost tracking
  • You want to enforce model access policies
  • You are already running LiteLLM or Bifrost in your infrastructure

​
LiteLLM

LiteLLM is an open-source proxy that supports 100+ models with a single OpenAI-compatible API.

​
Config

{
  providers: {
    litellm: {
      baseUrl: "http://localhost:4000",
      apiKey: "your-litellm-master-key",
    }
  },
  agents: {
    defaults: {
      model: "litellm/claude-sonnet-4-6",  // model as configured in LiteLLM
    }
  }
}
Or set environment variables: 
export LITELLM_API_BASE=http://localhost:4000
export LITELLM_API_KEY=your-master-key

​
Bifrost

Bifrost is WednesdayAI’s native LLM proxy integration. When requests are routed through Bifrost, the gateway captures the actual provider and model from x-bifrost-provider and x-bifrost-model response headers.

​
Config

{
  providers: {
    bifrost: {
      baseUrl: "http://bifrost.internal:8080",
      apiKey: "your-bifrost-key",
    }
  },
  agents: {
    defaults: {
      model: "bifrost/anthropic/claude-sonnet-4-6",
    }
  }
}

​
Header capture

When the @wednesdayai/bifrost extension is installed, the gateway reads x-bifrost-provider and x-bifrost-model from each response and records the actual provider used in run metadata. This is useful for auditing which provider handled each request. Install the Bifrost extension: 
openclaw plugins install @wednesdayai/bifrost

​
Troubleshooting

The proxy is not running. Check that LiteLLM or Bifrost is up and listening on the configured port.
The apiKey in your config does not match the proxy’s master key. Update providers.litellm.apiKey or providers.bifrost.apiKey.
The model ID must match exactly how it is configured in the proxy. Check your LiteLLM model list (litellm --list) or Bifrost admin UI.