DeepSeek-AI
DeepSeek-AI is a research organization that published a 2025 paper on large language models, cited in discussions of adaptive parallel reasoning and LLM scaling. Microsoft Research Asia – Singapore partners with Infocomm Media Development Authority (IMDA) LaMa inpainting model was used for training Doodle-in App SageMaker AI endpoint can be encrypted with AWS Key Management Service Oyindamola Omotuyi directs GPT-6 Astra Oyindamola Omotuyi uses NVIDIA Isaac Lab Oyindamola Omotuyi uses Newton Physics Engine Oyindamola Omotuyi uses Robotiq gripper Tae Kim used GPT-6 Astra Jensen Huang ANNOUNCED GTC Berlin keynote Thinking About Thinking authors come from Swiss AI Lab Argon solves RE-Bench Macquarie University LOCATED_IN Sydney President Trump speaks on Truth Social DeepSeek AI model USED_BY Snowflake DeepSeek AI model USED_BY Vercel DeepSeek AI model USED_BY OpenRouter DeepSeek AI model HAS_LOWER_DEVELOPMENT_COST_THAN OpenAI GPT-4 United States ordered federal civilian agencies to remove DeepSeek AI products Deep Agents harness uses glob· Graphiti 生成的实体摘要
关系变化
按生效时间- Deep AgentsUSES TOOLDeepSeek-AI
Deep Agents harness uses glob
- United StatesORDERED TO REMOVEDeepSeek-AI
United States ordered federal civilian agencies to remove DeepSeek AI products
- DeepSeek-AIUSED BYVercel
DeepSeek AI model USED_BY Vercel
- DeepSeek-AIUSED BYOpenRouter +
DeepSeek AI model USED_BY OpenRouter
- DeepSeek-AIUSED BYOffice of the Comptroller of the Currency
DeepSeek AI model USED_BY Snowflake
- DeepSeek-AIHAS LOWER DEVELOPMENT COST THANUnited States Justice Department
DeepSeek AI model HAS_LOWER_DEVELOPMENT_COST_THAN OpenAI GPT-4
- SageMaker AI endpointENCRYPTED WITHDeepSeek-AI
SageMaker AI endpoint can be encrypted with AWS Key Management Service
- Gemini 4 ArgonSOLVESDeepSeek-AI
Argon solves RE-Bench
- Macquarie UniversityLOCATED INDeepSeek-AI
Macquarie University LOCATED_IN Sydney
- Jensen HuangANNOUNCEDDeepSeek-AI
Jensen Huang ANNOUNCED GTC Berlin keynote
- DeepSeek-AIPACKAGESNVIDIA
Lenovo packages NVIDIA accelerated computing into locally operated AI factories for European cities.
- DeepSeek-AISPEAKS ONGoogle Labs
President Trump speaks on Truth Social
- DeepSeek-AIUSESNVIDIA Inception program
Oyindamola Omotuyi uses Robotiq gripper
- DeepSeek-AIUSESNVIDIA
Oyindamola Omotuyi uses Newton Physics Engine
- DeepSeek-AIDIRECTSGPT-6 Astra
Oyindamola Omotuyi directs GPT-6 Astra
- DeepSeek-AIUSED BYGPT-6 Astra
Tae Kim used GPT-6 Astra
- DeepSeek-AIUSESBSC AI Factory
Oyindamola Omotuyi uses NVIDIA Isaac Lab
- DeepSeek-AIUSED FOR TRAININGDoodle-in App
LaMa inpainting model was used for training Doodle-in App
- DeepSeek-AIADVISESSynthID
Pushmeet Kohli -> ADVISES -> SynthID Bio
- ESA 2010 sector S.14PARTNERS WITHDeepSeek-AI
Microsoft Research Asia – Singapore partners with Infocomm Media Development Authority (IMDA)
- Thinking About ThinkingAFFILIATED WITHDeepSeek-AI
Thinking About Thinking authors come from Swiss AI Lab
- DeepSeek-AIINITIATEDSynthID
Pushmeet Kohli -> INITIATED -> SynthID Bio
关系是 Graphiti 从报道中提取的事实陈述,不是原文引述;失效时间表示后来的报道显示这条关系已不成立。
相关事件 8
- LangChain博客发布AI智能体检索效率优化指南1 篇相关报道 · 2026-10-24
- 美国要求联邦机构移除DeepSeek AI1 篇相关报道 · 2026-10-20
- US bans DeepSeek AI from government use1 篇相关报道 · 2026-10-20
- Muse Spark Open发布与授权限制概览1 篇相关报道 · 2026-10-16
- Google Deepmind报告:自改进AI仍遥遥无期1 篇相关报道 · 2026-10-15
- NVIDIA承诺五年2.5亿美元支持英国AI安全研究1 篇相关报道 · 2026-10-14
- 欧洲城市部署 NVIDIA 驱动的 AI 智能体提升城市运营效率1 篇相关报道 · 2026-10-14
- Meta发布Muse Wonder和Wonder Full两款前沿AI模型1 篇相关报道 · 2026-10-13
来源报道 14
- AI 智能体检索效率优化:从工具选择到子代理过滤的完整指南LangChain Blog · 2026-10-24
- 美国要求联邦机构移除DeepSeek AITechCrunch · AI · 2026-10-20
- 美国禁用DeepSeek,政府系统清除中国AI模型BBC Business · 2026-10-20
- Muse Spark Open 现已上线 SageMaker JumpStart,提供两种量化方案和 1 亿 token 上下文窗口AWS Machine Learning Blog · 一手 · 2026-10-16
- Google 称自进化 AI 仍遥遥无期The Decoder · 2026-10-15
- Apate 潜入诈骗流程获取情报TechCrunch · AI · 2026-10-15
- NVIDIA宣布增投 UK 初创生态,涉及 AI Agent、生物模型及量子计算等多领域NVIDIA Blog · 一手 · 2026-10-14
- 欧洲城市部署 NVIDIA 功能强化的 AI 代理以运行更智能的城市运营NVIDIA Blog · 一手 · 2026-10-14
- Meta发布Muse Wonder和Muse Wonder Full两款前沿AI模型Decrypt · 2026-10-13
- NVIDIA 发布 Astra 与 Omniverse 库结合的开发者应用实例NVIDIA Blog · 一手 · 2026-10-09
- 用户用 ML-intern 训练七个小型模型,总成本仅103美元Hugging Face Blog · 一手 · 2026-10-08
- Introducing SynthID Bio: Google DeepMind's watermarking technology for AI-generated proteinsGoogle DeepMind · 一手 · 2026-09-30
- Microsoft Research Asia – Singapore 一周年:推动研究、合作与人才实现现实影响Microsoft Research · 一手 · 2026-09-29
- Import AI 469:Science AI、RSI 模拟器与 Zuk 的技术 pessimismImport AI · 2026-08-17