☸️EXO
Run frontier AI locally.
⭐ Very popular: 47k stars, gaining about 150 a week
View on GitHub ↗repo profile
momentum
durability
bus factor = how many people it takes to cover more than half the commits (6 months). 1 is a solo project; higher means the work is spread across a team. top-author share is the single busiest author's slice of those commits.
since we covered it
why it's a big deal
- Device-agnostic clustering: Works seamlessly across heterogeneous hardware without complex setup.
- Scalable model inference: Dynamic partitioning ensures models can be run across multiple devices, improving throughput.
- Easy integration: Offers a ChatGPT-compatible API for plug-and-play usage.
under the hood
- Built in Python under GPL-3.0 license.
- Supports a wide range of models (LLaMA, Mistral, Qwen, DeepSeek) and inference engines like MLX, tinygrad, and others.
- Zero-conf device discovery with peer-to-peer architecture; no central controller.
our take from PR#15, 2025-08-20
star history
- PR#15 30k 2025-08-20
- now 47k + 17k since first covered
curve is sampled from GitHub's star history, plus our own daily readings since we covered it; the dashed stretch is before we first covered it, the solid line since. figures at coverage are the numbers we printed then (approx.), current count is live.
understory
Better known than its recent output, coasting a little on attention.
- output, commits & releases
- clout, star velocity
output = commits & releases; clout = star velocity, both 0 to 100 monthly indices; the gap where output runs above clout is the understory. The understory →
covered in
-
DIY AI Cluster Framework
similar projects
compare these →- 🧰 Qwen3
Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud.
28k ACTIVE - 🧬 Second-Me
leaner, 16k stars
Train your AI self, amplify you, bridge the world
16k ACTIVE - 📡 Dask
leaner, 14k stars
Parallel computing with task scheduling
14k ACTIVE
comments
Sign in with GitHub to add your blip on EXO.