Ax LLM Framework

Ax is a powerful open-source TypeScript/JavaScript framework designed for building robust, production-grade LLM agents and AI pipelines. It draws inspiration from Stanford's DSPy to bring structured prompt optimization and multi-step reasoning to the Node.js ecosystem. Ax supports a wide range of AI providers through a unified API, making it easy to switch models or run experiments without rewriting your application logic. The framework emphasizes type safety, composability, and developer ergonomics for teams shipping AI features at scale.

• Multi-provider support: Connect to OpenAI, Anthropic, Google Gemini, Mistral, Cohere, and more through a single unified interface.
• DSPy-inspired prompt optimization: Use the built-in optimizer to automatically tune prompts and few-shot examples for better task performance.
• Typed signatures: Define input/output schemas with TypeScript types so the framework enforces structured LLM outputs at runtime.
• Agent & multi-agent orchestration: Build single agents or compose multiple agents into pipelines with tool use, memory, and handoffs.
• Streaming support: Handle streaming responses from LLMs natively for real-time user experiences.
• Function/tool calling: Easily define and register tools that agents can invoke, with automatic argument parsing and validation.
• Retrieval-Augmented Generation (RAG): Integrate vector stores and retrieval steps directly into agent pipelines.
• Observability: Built-in tracing and logging hooks to monitor agent behavior and debug production issues.
• Getting started: Install via npm (npm install @ax-llm/ax), set your provider API keys, define a typed signature, and run your first agent in minutes.