Manifest

An open-source LLM router that intelligently routes API requests to the most cost-effective AI model. Simple queries go to cheap models, complex ones go to powerful models — saving up to 70% on token costs with no code changes.

Access

After deployment, visit your domain to access the Manifest analytics dashboard. The API endpoint is available at the same domain for routing LLM requests.

What You Can Do After Deployment

1. Visit your domain — access the real-time analytics dashboard
2. Configure providers — add API keys for OpenAI, Anthropic, Google, and 300+ other models
3. Route requests — point your application to Manifest instead of the LLM provider directly
4. Monitor costs — view token usage, costs per model, and routing decisions in the dashboard
5. Set budgets — configure spending limits and model fallbacks
6. Analyze patterns — review which queries route to which models and optimize accordingly

Key Features

• Smart routing across 300+ LLM models from major providers
• Real-time analytics dashboard with cost tracking
• Token usage and spend monitoring per model
• Budget limits and model fallback chains
• Drop-in replacement for OpenAI API format
• Request scoring and intelligent model selection
• Full routing decision audit trail

License

MIT — GitHub