深度求索（DeepSeek），成立于2023年，专注于研究世界领先的通用人工智能底层模型与技术，挑战人工智能前沿性难题。基于自研训练框架、自建智算集群和万卡算力等资源，深度求索团队仅用半年时间便已发布并开源多个百亿级参数大模型，如DeepSeek-LLM通用大语言模型、DeepSeek-Coder代码大模型，并在2024年1月率先开源国内首个MoE大模型（DeepSeek-MoE），各大模型在公开评测榜单及真实样本外的泛化效果均有超越同级别模型的出色表现。和 DeepSeek AI 对话，轻松接入 API。
Public proof receipt / delayed sample
DeepSeek P1 · LAUNCH
DeepSeek updates its homepage to feature DeepSeek-V4 and DeepSeek-R1.
See the relevant before and after, when the change was found, and the official page behind it.
深度求索（DeepSeek），探索未至之境。专注于研究世界领先的通用人工智能底层模型与技术，开源 DeepSeek-V4、DeepSeek-R1 等前沿大模型。和 DeepSeek AI 对话，轻松接入 API。
Check the record
Read the evidence in plain language.
The before-and-after text above shows what was present on the official page at each capture. This summary explains the reviewed change.
DeepSeek updates its homepage to feature DeepSeek-V4 and DeepSeek-R1.
DeepSeek replaced its legacy company and model history with a current first-party description that explicitly features its open-source DeepSeek-V4 and DeepSeek-R1 frontier models.
- Before captured
- Aug 13, 2026, 4:43 AM PDT
- After captured
- Aug 13, 2026, 4:51 PM PDT
- Reviewed
- Aug 14, 2026, 1:36 PM PDT
WHY IT MAY MATTER
A messaging shift can reveal a new buyer, category claim, use case, or go-to-market emphasis.
Keep this record for the competitors that affect you.
Start with Scout for one competitor, then expand the watchlist when you need to.
Start Tracking →