Skip to content
latestai.eulatestai.eu
AT

AI Toolkit

by Ostris

Try itWebsiteDocsGitHub

Open-source GUI and CLI training suite for fine-tuning image and video diffusion models, including LoRAs, on consumer GPUs.

Available in EUPlatforms & InfraImage GenerationLibrary

Updated EU status checked Last shown

AI Toolkit at a glance

What is AI Toolkit?
AI Toolkit is a library from Ostris. Open-source GUI and CLI training suite for fine-tuning image and video diffusion models, including LoRAs, on consumer GPUs.
Who makes AI Toolkit?
AI Toolkit is made by Ostris.
Does AI Toolkit work in the EU?
Available in the EU. MIT-licensed code that can be downloaded and run locally anywhere, with no regional restrictions. Individual base models may carry their own licenses. Last verified 14 Nov 2026.
How much does AI Toolkit cost?
Open source, open source. Free, open-source software; users supply their own hardware or compute. Licence: MIT.
Where was AI Toolkit shown?
In 2 creator videos by AI Search. Most recently by AI Search on 20 Oct 2026, at 20:26 in “New top local AI image generator is here!”.

Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.

About

AI Toolkit, developed by Ostris, is a training suite for diffusion models. It can be run as a GUI or from the command line, and aims to support recent image and video models on consumer-grade hardware. The project is MIT-licensed and hosted on GitHub.

The README lists a long set of supported base models. Image models include FLUX.1, FLUX.2, Qwen-Image, Z-Image, HiDream, Chroma, Lumina2, SDXL, SD 1.5 and MOSS-Image. Instruction/edit models include FLUX.1 Kontext and Qwen-Image-Edit, and video models include the Wan 2.1 and 2.2 families. Edit-capable models are trained with control images in the dataset. A YouTube creator also notes that it works on macOS.

It is aimed at hobbyists and practitioners who want to train LoRAs or fine-tunes locally. Base-model weights are downloaded separately from Hugging Face and remain under their own licenses, which may impose restrictions independent of the toolkit's MIT license.

Key features
  • GUI and CLI interfaces
  • Trains LoRAs and fine-tunes for diffusion models
  • Supports many image, instruction/edit and video models (FLUX, Qwen-Image, Z-Image, Wan, etc.)
  • Designed for consumer-grade hardware
  • Edit-model training via control images in the dataset
  • Free and open source (MIT)
2 moments

As seen on YouTube

Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.

Keep exploring

Similar AI tools

Closest neighbours by meaning, not just by category.

PJ

PJiT

Model · Linum AI

1

2B-parameter pixel-space text-to-image diffusion model from Linum AI, trained from scratch for about $21K; Apache 2.0.

Available in EUImage Generation
Release date n/aOpen source
K2

Krea 2 LoRAs

App · Krea AI

1

LoRA training on Krea's platform to teach Krea 2 a custom style, character, face or product from a few images.

Available in EUImage Generation
Release date n/aFreemium
LT

LTX Trainer

Library · Lightricks

1

Official Lightricks package for training and fine-tuning LTX-2 video models, including LoRAs for characters, effects and styles.

Available in EUPlatforms & Infra
Release date n/aOpen source
I1

i1

Model · Princeton

1

Fully open 3B text-to-image diffusion model from Princeton with released weights, training code, and data pipeline.

Available in EUImage Generation
Release date n/aOpen source
SE

SeedVR2

Model · ByteDance

1

ByteDance's open-weights one-step diffusion transformer for restoring and upscaling video and images.

Available in EUImage Generation
10 Jun 2025Open source
MI

MOSS Image 8-step DMD LoRA

Model

1

8-step distillation LoRA for MOSS-Image-SFT that runs without CFG, for ComfyUI.

Available in EUImage Generation
16 Oct 2026Open source