Overview
News
Salaries
Products
People
Growth
Offices
Financials

Funding and backers

Backers
2
Y Combinator, Lightspeed Venture Partners
Momentum
0/100
no fresh signals
Last sign of life
1 mo ago
a posting or a news item
The story so fareach stage dated by its source: the company's releases, the press, registers, filings
Jun 22 On the Hacker News front page GLM-5.2 – How to Run Locally · 617 points Mar 17 On the Hacker News front page Unsloth Studio · 388 points Jan 28, 2025 On the Hacker News front page Run DeepSeek R1 Dynamic 1.58-bit · 767 points Jun 2024 Went through Y Combinator 2023 Founded
Tractionpublic signals, read monthly
Web presence
top 0.02%
of 133.2M domains by links, Common Crawl
Hacker News stories
75
best: 767 points
Backed byfrom the funds' portfolios and accelerator catalogs

Overview

Unsloth AI builds open source software that makes fine tuning and training large language models dramatically faster and less memory hungry. Its kernels and quantisation work let researchers train sizeable models on a single consumer graphics card, and its notebooks are among the most used entry points for practical fine tuning. The company maintains the library openly while offering a commercial version for teams that need scale and support; its library is downloaded millions of times and integrated into popular training stacks.

News

News sohu.com 2 months ago
Tongyi Qianwen官方点赞Unsloth Dynamic V3让Qwen3.8-27B更小更聪明1-bit版本仅需8GB内存即可运行
对本地AI生态的意义Qwen3.8-27B本身具备原生多模态能力拥有262K上下文窗口可扩展至约100万tokens在编码和办公工作流等场景中表现优异并采用Apache 2.0开源许可结合Uns…
Read more
Report
News iThome 2 months ago
Unsloth更新Dynamic v3.0 GGUF量化方法相同檔案尺寸下提高量化模型精度
Dynamic v3.0針對GGUF量化改用更多元校準資料並調整模型層量化策略Unsloth測試顯示相同檔案尺寸下可降低量化造成的精度損失另加入連續32個詞元輸出比較觀察量化模型是否逐漸偏離BF16模型
Read more
Report
News sohu.com 2 months ago
通义千问官方点赞Unsloth Dynamic V3 让 Qwen3.8-27B 更小更聪明1-bit 版 8GB 内存就能跑
通义千问Qwen官方近日公开感谢 Unsloth 团队对其发布的 Qwen3.8-27B GGUF量化版本表示高度认可并称该版本
Read more
Report
News huggingface 2 months ago
unsloth/Qwen3.8-27B-GGUF Introducing Unsloth Dynamic v3 Qwen3.8
Hey everyone! We're releasing new Qwen3.8-27B GGUFs with 10% higher accuracy for the same size. This uses a new version of Dynamic v3.0
Read more
Report
News vc.ru 2 months ago
Пляски с CUDA можно заканчивать. Unsloth упаковал обучение 500+ нейросетей в бесплатное приложение - AI на vc.ru
Unsloth Studio, обучение нейросетей, 500+ моделей, локальные модели, визуальный интерфейс, экспорт в GGUF, работа с файлами, NVIDIA GPU
Read more
Report
News huggingface 2 months ago
unsloth/Qwen3.8-27B-GGUF Hugging Face
We're on a journey to advance and democratize artificial intelligence through open source and open science.
Read more
Report
News unsloth.ai 2 months ago
Qwen3.8 - How to Run Locally Unsloth Documentation
Guide to running Qwen3.8 quants including Qwen3.8-27B on your local setup.
Read more
Report
News huggingface 2 months ago
unsloth/DeepSeek-V4-Pro-0813-GGUF Hugging Face
We're on a journey to advance and democratize artificial intelligence through open source and open science.
Read more
Report
Show more news

Offices

Where the company hires and what each office is for
Cities
1
Hiring now
0
Countries
1
San Francisco United States
HQ