# ax-llm/ax

The pretty much "official" DSPy framework for Typescript

Repository: https://github.com/ax-llm/ax
Canonical: https://ross.abutalabs.com/products/ax
Homepage: http://axllm.dev
Language: TypeScript
License: Apache-2.0
License Family: permissive
Topics: ai, cohere, llm, openai, typescript, javascript, nodejs, claude, large-language-models, opensource, anthropic, gemini, google, gpt-4, ollama, rag, vectordb, google-gemini, dspy, webllm
Last push: 2026-08-25T20:25:04+00:00

## Health v2 (maintenance only)
Score: 93/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 91
- inputs: {"age_days": 1287, "days_push": 8, "days_rel": 8, "gap_med": 0.0, "n_releases_24m": 413}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2885, forks 190 (observed 2026-08-28T04:07:28.158718+00:00)

## What it is
Ax is a TypeScript-first LLM programming framework implementing the DSPy model, providing typed signatures, agents, flows, and optimizers like GEPA across many LLM providers. The same semantic core is also compiled into Python, Java, C++, Go, and Rust libraries.

## Use cases
- build typed structured LLM calls with validated outputs
- create DSPy-style prompt programs in TypeScript
- build LLM agents with memory and tool use
- optimize prompts and few-shot demos with GEPA
- build multi-step LLM workflows with branching and loops
- switch between OpenAI, Anthropic, Gemini, and local models
- build RAG pipelines with vector databases

## When to choose
- you want DSPy-style prompt programming in TypeScript or Node.js
- you need typed, validated structured outputs from LLMs
- you want one framework spanning multiple LLM providers
- you need prompt optimization with metrics and teacher/student models
- you want the same LLM program model across several languages

## When to avoid
- you only need raw chat completions without structure or optimization
- you need a language not covered by the compiled libraries
- you prefer Python-native DSPy with its full ecosystem
- you want a no-code or UI-driven LLM builder

## Facets
- artifact type: framework
- maturity: active
- function: llm-inference, agent-framework, rag, prompt-engineering, sdk, machine-learning
- domain: large-language-models, artificial-intelligence, developer-tools
- platform: python, jvm, go, rust, cpp, cross-platform
- tags: dspy, typed-signatures, llm-optimization, gepa, multi-provider, structured-output, llm-workflows, ai-agents, retrieval-augmented-generation, nodejs, typescript

## Member repositories
- ax-llm/ax (main) score 93

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:28.158718+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:48:18.697232+00:00, confidence not recorded.
  - readme: https://github.com/ax-llm/ax (fetched 2026-08-28T04:07:28.158718+00:00, sha 84f46c08a3c8)
  - homepage: http://axllm.dev (fetched 2026-08-29T09:51:07.417059+00:00, sha 43baf3cf8940)
  - site_page: http://axllm.dev/typescript/api/ai (fetched 2026-08-29T09:51:07.420446+00:00, sha a20e47b7055f)
  - site_page: http://axllm.dev/typescript/skills (fetched 2026-08-29T09:51:07.422555+00:00, sha 3d4999a6a367)
  - site_page: http://axllm.dev/typescript/concepts/optimization (fetched 2026-08-29T09:51:07.424376+00:00, sha ff47f087449f)
  - site_page: http://axllm.dev/typescript/concepts/telemetry (fetched 2026-08-29T09:51:07.426453+00:00, sha 99bd7645d3ff)
- Data as of 2026-08-30T08:39:29.467469+00:00.