Research & News

Dedicated to exploring the essence of AGI — driven by curiosity, answering the biggest questions with the longest-term vision.

Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient.

NewsSeptember 10, 2026

Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient.

Introducing the smallest model in our new architecture family, with native visual understanding. Designed for greater capability, faster inference, and higher throughput.

NewsApril 24, 2026

DeepSeek-V4 Preview: Entering the Era of Affordable Million-Token Context

DeepSeek-V4 Preview is officially live & open-sourced. Welcome to the era of cost-effective 1M context length.

NewsDecember 1, 2025

DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models

DeepSeek-V3.2 and V3.2-Speciale launch with world-leading reasoning, thinking in tool-use, and gold-medal performance in IMO, CMO, ICPC & IOI 2025.

NewsSeptember 29, 2025

Introducing DeepSeek-V3.2-Exp

DeepSeek-V3.2-Exp debuts DeepSeek Sparse Attention (DSA) for faster, more efficient training & inference on long context, with API prices cut by 50%+.

NewsSeptember 22, 2025

DeepSeek-V3.1 is now DeepSeek-V3.1-Terminus

DeepSeek-V3.1-Terminus improves language consistency, reduces CN/EN mix-ups, and upgrades Code Agent & Search Agent performance.

Research Index

June 24, 2026DeepSeek-V4: Towards Highly Efficient Million-Token Context IntelligenceFebruary 25, 2026DualPath: Breaking the Storage Bandwidth Bottleneck in Agentic LLM InferenceJanuary 28, 2026DeepSeek-OCR 2: Visual Causal FlowJanuary 12, 2026Engram: Conditional Memory via Scalable LookupDecember 31, 2025mHC: Manifold-Constrained Hyper-ConnectionsDecember 2, 2025DeepSeek-V3.2: Pushing the Frontier of Open LLMsNovember 27, 2025DeepSeekMath-V2: Towards Self-Verifiable Mathematical ReasoningNovember 1, 2025Linear-Programming-Based Load Balancer (LPLB)October 21, 2025DeepSeek-OCR: Contexts Optical CompressionMay 14, 2025Insights into DeepSeek-V3: Scaling Challenges & Hardware Reflections