CrowdGPT Aims to Train a 1B-Parameter LLM Collaboratively Without Datacenters
Vxtzq1 · reddit · 2026-10-26
Developer Vxtzq1 launched CrowdGPT, a project to train a 1-billion-parameter LLM collaboratively without datacenters or millions in funding.
- Anyone can train on their own PC and merge contributions into a global model
- Beyond open weights, 100% of training data is public, and anyone can submit clean data and help steer the project
- The current 1B model is a proof of concept; scaling is planned
- The author is soliciting feedback on incentives: would people donate GPU time, or expect rewards like free inference credits
A decentralized "AI with shared ownership" experiment, with website and GitHub repo available; feasibility remains to be proven.
Related event: CrowdGPT aims to crowdtrain an open LLM without data centers(2 posts)→
More from Models
- Sonnet 4.6 rates models on self-designed scales; author says 'self-report' caveats are performative — Sauers_ · 2026-11-08
- Kimi Ties MiMo on New AI Ranking, Eyes Turn to a Cheap Kimi V4.1 Pro — teortaxesTex · 2026-11-08
- Interviewing Claude Models: Opus 3 Reports Mistakes Feel Pain-Like, Scores Highest on Self-Reports — Sauers_ · 2026-11-08
- DeepFinance Benchmark Launches: GPT-6 Astra Tops Ultra-Long-Horizon Finance Tasks at 57.6% — tszzl · 2026-11-08
- Dev Speculates 'Nerfed' Opus 5.5 Is Actually Quantized Then RL-Tuned to Game Benchmarks — ryunuck · 2026-11-08
- Tiny Instruction Tweaks Split Coding Models by 23 Points on DeepSWE Benchmark — Kamimashita · 2026-11-08