Manifest

Manifest is an open-source LLM router that sits between your application and multiple language models. It inspects each request and routes it to the most suitable model based on rules you define—such as cost limits, latency targets, or quality requirements—while handling fallbacks and retries automatically.

Manifest lets you route routine or low-risk traffic to cheaper models and reserve expensive, high-quality models only for requests that truly need them. By tuning routing strategies and monitoring usage, teams often significantly reduce overall token spend without changing their product UX.

In most cases, you only need to point your LLM calls to Manifest's API endpoint instead of directly to a single provider. Your routing logic, provider configs, and model choices are then centralized in Manifest, which minimizes changes to application code.

Yes. Manifest is released as an open-source project, so you can inspect the code, run it on your own infrastructure, and customize it to fit your needs without licensing fees. You only pay for the underlying LLM providers you choose to use.

Manifest is designed to be vendor-agnostic and can work with a wide range of commercial and open-source LLMs exposed via API. You can configure multiple providers, add new models as they become available, and swap them out without changing your client code.