Model ReleasesDeepSeek AI releases R1, a 671B MoE reasoning model that rivals o1. Open source, RL-trained, and causing market shockwaves.almost 2 years ago6 min read
