Featured on
Foundation models · Hangzhou, China
DeepSeek's R1 model forced every major AI lab to publicly defend its compute spending in early 2025, and remains the clearest evidence that frontier-level AI research is no longer a US-only contest.
DeepSeek is a Hangzhou-based AI research lab founded in 2023 by Liang Wenfeng, who built the company's computing infrastructure through High-Flyer, the quantitative hedge fund he co-founded. DeepSeek's open-weight model releases - particularly its R1 reasoning model - triggered a sharp reassessment in early 2025 of how much compute frontier-level AI performance actually requires, briefly moving US technology-stock markets on the news.
The company has continued to release open-weight models through 2026 and is reported to be preparing an IPO on the Chinese mainland, with a filing potentially as soon as this year and a debut targeted for 2027 - a notable move for a company that spent its first two years avoiding the spotlight entirely.
DeepSeek-R1
Open-weight reasoning model whose efficiency claims reshaped industry compute assumptions in early 2025.
DeepSeek-V3
Open-weight general-purpose large language model.
The maker of ChatGPT and the GPT model family, and the company most responsible for the current wave of consumer generative AI.
AI safety-focused lab behind the Claude model family, founded by former OpenAI researchers.
Alphabet's unified AI research and product lab, formed from the 2023 merger of DeepMind and Google Brain, and the maker of the Gemini model family.
