How we packed 2500 people in a hockey arena for the world's greatest demo day

Today we’re launching INTELLECT-2: The first decentralized 32B-parameter RL training run open to join for anyone with compute — fully permissionless. Scaling towards frontier reasoning across coding, math and science.

“decentralized training will alter the political economy of superintelligence” - cofounder of anthropic

The Metamorphosis of Prime Intellect

synthetic-2 soon planetary-scale decentralized inference lots of contributor slots get ur 3090s ready

Releasing INTELLECT-2: We’re open-sourcing the first 32B parameter model trained via globally distributed reinforcement learning: • Detailed Technical Report • INTELLECT-2 model checkpoint primeintellect.ai/blog/intellect…

one step closer to Prime Intellect

all your idle compute are belong to us

real

prime intellect masterplan

we may or may not be hosting this at a church

launch in two hours

Launching SYNTHETIC-2: our next-gen open reasoning dataset and planetary-scale synthetic data generation run. Powered by our P2P inference stack and DeepSeek-R1-0528, it verifies traces for the hardest RL tasks. Contribute towards AGI via open, permissionless compute.

still wild to me how china is completely monopolizing open-source ai right now every other day there's a new frontier release from bytedance, baidu, xiaomi, deepseek, qwen, tencent, huawei, moonshot, zhipu, minimax, stepfun, … we really need to accelerate here