DeepSeek is a Chinese AI research lab founded in 2023 as a subsidiary of High-Flyer, a quantitative trading firm based in Hangzhou. The company gained international attention in January 2025 with DeepSeek R1, an open-weight reasoning model that rivaled OpenAI's o1, and DeepSeek V3, which achieved GPT-4o-level performance at a reported training cost of just $5.6 million. DeepSeek releases all models under MIT license and has been influential in demonstrating that frontier-level AI can be developed at dramatically lower cost than previously assumed.
Key Metrics
Headcount
Facts
13Related Wiki Pages
Top Related Pages
DeepSeek Models
DeepSeek V3
DeepSeek R1
AI Safety Multi-Actor Strategic Landscape
Analyzes how AI risk varies by which actors develop TAI, finding actor identity may determine 40-60% of total risk variance. Updated with 2025-2026...
AI Misuse Risk Cruxes
Key uncertainties that determine views on AI misuse risks, including capability uplift, offense-defense balance, agentic misuse vectors, open-weigh...