# ostris/ai-toolkit

The ultimate training toolkit for finetuning diffusion models

Repository: https://github.com/ostris/ai-toolkit
Canonical: https://ross.abutalabs.com/products/ai-toolkit
Language: Python
License: MIT
License Family: permissive
Last push: 2026-08-26T17:06:37+00:00

## Health v2 (maintenance only)
Score: 73/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 35, longevity 82
- inputs: {"age_days": 1155, "days_push": 7, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 11838, forks 1500 (observed 2026-08-28T04:10:50.152060+00:00)

## What it is
An all-in-one open-source training toolkit for finetuning diffusion models (image and video) on consumer-grade hardware. It supports many recent models like FLUX, SDXL, and Qwen-Image, and can be run via a GUI or CLI.

## Use cases
- train a LoRA for FLUX.1 on my own images
- finetune stable diffusion xl on consumer gpu
- train a custom image model without writing code
- fine-tune a diffusion model with a gui
- train video diffusion models locally
- create custom ai image model from photos
- run diffusion model training from the command line

## When to choose
- you want to finetune or train LoRAs for supported diffusion models on consumer hardware
- you want a GUI plus CLI for training image or video diffusion models
- you want broad support for the latest open image models like FLUX, SDXL, and Qwen-Image

## When to avoid
- you need to train LLMs or non-diffusion model architectures
- you only want to run inference on already-trained models
- you need a managed cloud training service rather than a local toolkit

## Facets
- artifact type: application
- maturity: active
- function: machine-learning, llm-training, cli, gui
- domain: machine-learning, image-processing, deep-learning, artificial-intelligence
- platform: python, windows
- tags: diffusion-models, fine-tuning, lora, flux, stable-diffusion, training-toolkit, image-generation, video-models, linux, macos, gpu

## Member repositories
- ostris/ai-toolkit (main) score 73

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:50.152060+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 2026-08-29T17:15:17.105169+00:00, confidence not recorded.
  - readme: https://github.com/ostris/ai-toolkit (fetched 2026-08-28T04:10:50.152060+00:00, sha 45b10c89da9a)
- Data as of 2026-08-30T08:39:29.467469+00:00.