# AI-Hypercomputer/maxtext

A simple, performant, and scalable Jax LLM!

Repository: https://github.com/AI-Hypercomputer/maxtext
Canonical: https://ross.abutalabs.com/products/maxtext
Homepage: https://maxtext.readthedocs.io
Language: Python
License: Apache-2.0
License Family: permissive
Topics: large-language-models, llm, gpt, deepseek, fine-tuning, gemma2, gemma3, jax, llama2, llama3, llama4, mistral, mixtral, sft
Last push: 2026-08-26T21:47:23+00:00

## Health v2 (maintenance only)
Score: 97/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 98, longevity 91
- inputs: {"age_days": 1282, "days_push": 7, "days_rel": 12, "gap_med": 18, "n_releases_24m": 20}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2408, forks 592 (observed 2026-08-28T04:06:44.211945+00:00)

## What it is
MaxText is a high-performance, scalable open-source LLM training library written in pure Python/JAX, targeting Google Cloud TPUs and GPUs. It provides reference implementations of models like Gemma, Llama, DeepSeek, Qwen, and Mistral with support for pre-training at massive scale and post-training techniques such as SFT, GRPO, and GSPO.

## Use cases
- pre-train large language models on TPU clusters
- fine-tune Llama or Gemma with supervised fine-tuning
- run reinforcement learning post-training like GRPO for LLMs
- scale LLM training to tens of thousands of chips
- train DeepSeek or Mistral models in JAX
- benchmark high MFU LLM training on GPUs and TPUs

## When to choose
- you need to train or fine-tune LLMs on Google Cloud TPUs or GPUs at large scale
- you want a pure Python/JAX codebase that is easy to fork and modify
- you need high hardware utilization without hand-written kernel optimizations
- you want reference implementations of popular open model architectures

## When to avoid
- you need PyTorch-based training or ecosystem compatibility
- you only need LLM inference rather than training
- you lack access to TPU or GPU hardware
- you need diffusion or image generation models (see MaxDiffusion instead)

## Facets
- artifact type: library
- maturity: active
- function: llm-training, machine-learning, deep-learning
- domain: large-language-models, deep-learning, machine-learning
- platform: python, cloud
- tags: jax, tpu, fine-tuning, pre-training, sft, grpo, gemma, llama, deepseek, mistral, xla, gpu

## Member repositories
- AI-Hypercomputer/maxtext (main) score 97

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:44.211945+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 2026-08-30T02:34:02.334993+00:00, confidence not recorded.
  - readme: https://github.com/AI-Hypercomputer/maxtext (fetched 2026-08-28T04:06:44.211945+00:00, sha 120db8c0fb0e)
  - registry_pypi: https://pypi.org/pypi/maxtext/json (fetched 2026-08-29T10:14:41.709723+00:00, sha 4e4a2656cf14)
- Data as of 2026-08-30T08:39:29.467469+00:00.