Unsloth
Open-source framework and desktop app for running and fine-tuning open text, vision, audio and diffusion models on a user's own hardware, with LoRA, QLoRA, full fine-tuning and reinforcement-learning support and an OpenAI-compatible local API.
Three plans on unsloth.ai/pricing: Free ('Freeware of our standard version of unsloth', Apache-2.0 per the GitHub API), 'unsloth Pro' ('2.5x faster training + 20% less VRAM', Contact us) and 'unsloth Enterprise' ('Unlock 30x faster training + multi-node support + 30% accuracy', Contact us). Neither paid plan publishes a rate. Pro and Enterprise are tiers of this product, not separate products.
What it does
- Model training
- Model inference
Sources
- Pricing
-
Read in a real browser 2026-09-18; plan names and copy quoted above. Licence confirmed Apache-2.0 via api.github.com/repos/unslothai/unsloth (76,318 stars, last pushed 2026-09-18).
- Description
-
Official documentation · 18 Sep 2026
Read in a real browser 2026-09-18 (unsloth.ai returns HTTP 403 to a plain fetch). The docs' own words: 'Unsloth is an open-source framework for running and training LLMs' and 'Unsloth lets you run and train AI models on your own local hardware via an open-source UI.'
- Sold within
-
Bought on its own; the free tier is the open-source package and the paid tiers are licensed directly.
- URL
-
Official documentation · 18 Sep 2026
Homepage read in a real browser 2026-09-18: 'Meet Unsloth Desktop — The first desktop app to run and train AI models. Open-source. Free. 100% Local.'
Something wrong here? Send a correction — quote this product id: unsloth.