DeepSeek AI
To solve the challenge of AGI (Artificial General Intelligence) with high efficiency and lower costs, specializing in open-source LLMs.
At a Glance
- Developers
- AI Researchers
- Enterprise Software Companies
- Tech Startups
AI Tools by DeepSeek AI
(1)
Awesome DeepSeek Integration
DeepSeek Integrations Directory
Discussions
No discussions yet
Be the first to start a discussion about DeepSeek AI
Latest News
Launching DeepSeek-V3.2 & DeepSeek-V3.2-Speciale — Reasoning-first models built for agents!
DeepSeek-R1 Release: Performance on par with OpenAI-o1
DeepSeek shocks the market with $6M model, causing global tech shifts
DeepSeek-V3.1 update released, merging V3 and R1 strengths
Products & Services
The latest flagship large language model with enhanced agent capabilities and reasoning modes (thinking and non-thinking).
A reasoning-focused model using reinforcement learning, competitive with the world's most advanced reasoning AI.
An open-source code language model that outperforms GPT-4-Turbo in coding benchmarks.
A low-cost, high-throughput API platform for developers to integrate DeepSeek models into applications.
Market Position
The high-quality, ultra-low-cost open-source challenger to OpenAI and Anthropic, emphasizing transparency and efficiency.
Leadership
Founders
Liang Wenfeng
Co-founder of High-Flyer Quant (a top Chinese hedge fund). Graduate of Zhejiang University with a background in engineering and quantitative trading. Known for technical leadership in AI architecture.
Li Huan
Core partner and major shareholder in Hangzhou DeepSeek and High-Flyer Quant. Deeply involved in the strategic direction of the company's AI research.
Executive Team
Liang Wenfeng
CEO & Founder
Founder of High-Flyer Quant; AI researcher and quantitative finance pioneer.
Li Huan
Core Partner / Director
Major stakeholder and strategic leader from the High-Flyer ecosystem.
Board of Directors
Founding Story
Spun off from High-Flyer Quant, a leading Chinese hedge fund. Initially developed as an internal deep-learning unit (Fire-Flyer) for quantitative trading, it was transformed into a dedicated AI lab to focus on foundational LLM research and open-source models.
Business Model
Revenue Model
API-based usage (token consumption), mobile app subscriptions (VIP), and enterprise solutions.
Pricing Tiers
Price for Input (Cache Miss). Input (Cache Hit) is $0.028/1M tokens. Output is $0.42/1M tokens.
Access to chat and basic features.
Priority access and advanced features in the mobile application.
Target Markets
- Developers
- AI Researchers
- Enterprise Software Companies
- Tech Startups
- Software Engineering / Coding
- Mathematical Research
- Enterprise AI Agents
- General Purpose Conversational AI
- ByteDance
- Tencent
- Groq
- Nebius