Pluralis

ProjectWatch17

Decentralized training for open foundation models

InfrastructureAIDecentralized Computing

About

Pluralis is building decentralized infrastructure for training foundation models across non-collocated consumer GPUs. Their Agora pretraining stack achieves near-centralized efficiency without NVLink or InfiniBand, and their research on gossip-based distributed training and architecture warm-up addresses key stability and robustness challenges in decentralized AI.

Social Signals

Hotwatch17/100

X Followers10.9K

Trust ScoreCalculating…

Recent News

Decentralized AI training networks like @Pluralis' Agora are the only true counterbalance to centralized AI. https://x.com/Pluralis/status/2065899910080115117

@jbrukh◆5 sourcesJun 13, 2026

Whole section on Pluralis in Chamath's substack this week right under details of Anthropic's monster round.

@jbrukh◆Jun 8, 2026

As of today, there’s essentially 5 companies who have successfully completed meaningful SOTA moving decentralized *pretraining* runs: - @PrimeIntellect (10B INTELLECT-1, Oct 24) - @Pluralis (7.5B Node0, Oct 25; 8B Agora launch, May 26) - @NousResearch (40B Consilience, May 25) - @covenant_ai (Coven

@coinfund◆Jun 4, 2026

There are other teams talking a big talk with respect to training, and a select few like iota that have pushed some real research in the space, but in this space theres no team close to what Pluralis has put out. Unfortunately ive been one shot by the idea of distributed trainig and have spent way

@kelxyz_◆Jun 3, 2026