DeepSeek-R1
DeepSeek · · Open weights
DeepSeek released an open-source reasoning model matching OpenAI o1's benchmark performance, under MIT licence, at a fraction of the training cost. NVIDIA lost $589B in market cap on the following trading day.
Why it matters
First open-source reasoning model to match proprietary performance. Triggered a global reassessment of AI economics. Reasoning capability was no longer a proprietary moat.
This is LLM World's historical reading, not a claim by DeepSeek. The facts above and the sources below are what it rests on.
The model record for DeepSeek-R1 →
Sources
Record status: verified · confidence 85/100
DeepSeek's own announcement, dated January 20, 2025. The arXiv paper was submitted January 22, 2025. The widely used date is January 20.
- deepseek.com Official announcement archived
- arxiv.org Official documentation archived