One Interface.
Any LLM Provider.

any-llm is a library that provides a single interface to various Large Language Model (LLM) providers, allowing you to adapt to changing needs, markets, or costs, whether you're building agents or any other LLM-powered application. Switch providers in seconds without rewriting your code.

Why any-llm?

Today’s LLM ecosystem is fragmented. Each provider has unique APIs, parameters, response formats, and feature sets. any-llm unifies them under a simple, single API so you can adapt to changing needs, markets, or costs without code rewrites.

Whether you’re using LLMs hosted by AWS, Azure, OpenAI, Anthropic, Mistral, or others, any-llm makes it easy to switch models and providers on demand.

Illustration of a window with gears

One Interface,
Any Provider

Call multiple LLMs and providers with a single function signature. No provider-specific wrappers to maintain.

Illustration of hammer and wrench with a checkmark

Built with Official SDKs

Uses official SDKs when available to ensure compatibility, performance, and stability.

Illustration of a cloud

No Extra Infrastructure Needed

No proxies. No gateways. Just plug in your provider keys and start building.

Illustration of a robot head

Used in the Mozilla.ai Agent Platform

Every agent built on our platform runs on any-llm, enabling seamless provider switching as your needs evolve.

Simplified Management and Tracking for all your LLM Providers.

Built on top of any-llm, this platform allows you to securely manage your provider API keys for both cloud and local models.
Monitor usage, uncover trends, and confidently scale your AI applications with client-side encryption.
Privacy-first design. No lock-in. No new SDKs.

Now in Beta and free to use.

Get Started
with any-llm

Try it yourself and switch LLM providers without changing your code.