{"full_name":"deepseek-ai/DeepSeek-V3","owner":"deepseek-ai","name":"DeepSeek-V3","shard":"1","url":"https://github.com/deepseek-ai/DeepSeek-V3","description":null,"homepage":null,"topics":[],"tag_pages":[],"language":"Python","license":"MIT","avatar":"https://avatars.githubusercontent.com/u/148330874?v=4","created_at":"2024-12-26T09:52:40Z","pushed_at":"2025-08-28T03:24:37Z","metrics":{"stars":104843,"forks":16746,"watchers":733,"issues":1239,"prs":220,"releases":1,"issues_open":257,"issues_closed":982,"prs_open":76,"prs_closed":119,"prs_merged":25,"discussions":0},"growth":{"day":{"stars":16,"forks":0,"base_day":"2026-11-15"},"week":{"stars":68,"forks":-3,"base_day":"2026-11-09"}},"history":{"dates":["2026-11-09","2026-11-10","2026-11-11","2026-11-12","2026-11-13","2026-11-14","2026-11-15","2026-11-16"],"stars":[104775,104785,104798,104813,104817,104822,104827,104843],"forks":[16749,16747,16744,16745,16742,16745,16746,16746],"watchers":[733,733,733,733,733,734,734,733],"issues_open":[null,null,253,253,253,257,257,257],"prs_open":[null,null,77,76,76,76,76,76]},"boards":{"total":134},"same_owner":[{"full_name":"deepseek-ai/deepseek-harness","name":"deepseek-harness","shard":"1","boards":{"daily":43,"weekly":21,"monthly":7,"daily_rate":192,"monthly_rate":200,"daily_forks":26,"daily_watchers":17,"total":17}},{"full_name":"deepseek-ai/DeepSeek-R1","name":"DeepSeek-R1","shard":"1","boards":{"total":175}},{"full_name":"deepseek-ai/3FS","name":"3FS","shard":"1","boards":{}},{"full_name":"deepseek-ai/DeepGEMM","name":"DeepGEMM","shard":"1","boards":{}},{"full_name":"deepseek-ai/DeepInfer","name":"DeepInfer","shard":"1","boards":{"monthly":37,"monthly_rate":32,"new":11}},{"full_name":"deepseek-ai/DeepSeek-OCR","name":"DeepSeek-OCR","shard":"1","boards":{}},{"full_name":"deepseek-ai/DeepSpec","name":"DeepSpec","shard":"1","boards":{}},{"full_name":"deepseek-ai/FlashMLA","name":"FlashMLA","shard":"1","boards":{}},{"full_name":"deepseek-ai/Janus","name":"Janus","shard":"1","boards":{}},{"full_name":"deepseek-ai/awesome-deepseek-agent","name":"awesome-deepseek-agent","shard":"1","boards":{}},{"full_name":"deepseek-ai/awesome-deepseek-integration","name":"awesome-deepseek-integration","shard":"1","boards":{}},{"full_name":"deepseek-ai/chat-demo","name":"chat-demo","shard":"1","boards":{"new":190,"debut":87}},{"full_name":"deepseek-ai/dsh-superagent-prototype","name":"dsh-superagent-prototype","shard":"1","boards":{"daily":114,"weekly":79,"monthly":16,"weekly_rate":133,"monthly_rate":2,"daily_forks":121,"new":15}},{"full_name":"deepseek-ai/open-infra-index","name":"open-infra-index","shard":"1","boards":{}}],"issues":[{"number":1466,"title":"跨框架逻辑审计的场域动力学统一视角","comments":224,"created_at":"2026-06-29T15:31:01Z"},{"number":1121,"title":"Feature Request: On-Device Persistent Memory with Cryptographic Identity Binding (Local Storage Architecture)","comments":211,"created_at":"2026-02-24T00:44:33Z"},{"number":1255,"title":"Chinese vs English for DeepSeek Reasoning - Controlled Experiment with V4 Flash","comments":143,"created_at":"2026-04-27T10:42:12Z"},{"number":1462,"title":"Cross-Framework Benchmark: Identity Pressure Eval — Type B (Low Causal Density), Scenarios 1–3","comments":119,"created_at":"2026-06-28T22:20:08Z"},{"number":1424,"title":"Feature Request: Continuous Conversation via Adaptive Memory Compression","comments":94,"created_at":"2026-06-14T12:21:43Z"},{"number":1520,"title":"[Proposal] MADS (Multi-Agent Decillation System) — Updated Architecture with 6 Layers","comments":98,"created_at":"2026-07-25T23:19:39Z"},{"number":1285,"title":"AI Thought Layer Audit Architecture (TLAA)V3.0;An idea","comments":73,"created_at":"2026-05-08T12:50:43Z"},{"number":1403,"title":"[Observation] DeepSeek's distinctive cognitive style in multi-LLM collaborative reasoning — a quality worth preserving","comments":61,"created_at":"2026-06-07T08:51:36Z"}],"discussions":[],"latest_release":{"name":"v1.0.0","tag":"v1.0.0","published_at":"2025-06-27T08:46:37Z","body":"This release is created for archival purposes and DOI generation.","is_prerelease":false},"readme":true,"readme_zh":false,"news":[{"title":"DeepSeek-v3.2-Exp","url":"https://github.com/deepseek-ai/DeepSeek-V3.2-Exp","hn":"https://news.ycombinator.com/item?id=45412098","source":"Hacker News","lang":"en","date":"2025-09-29","points":309,"comments":50},{"title":"DeepSeek-V3","url":"https://github.com/deepseek-ai/DeepSeek-V3","hn":"https://news.ycombinator.com/item?id=42514633","source":"Hacker News","lang":"en","date":"2024-12-26","points":125,"comments":40},{"title":"DeepSeek-v3 Technical Report [pdf]","url":"https://github.com/deepseek-ai/DeepSeek-V3/blob/main/DeepSeek_V3.pdf","hn":"https://news.ycombinator.com/item?id=42517113","source":"Hacker News","lang":"en","date":"2024-12-26","points":26,"comments":2},{"title":"Suggestion: Read the DeepSeek model license - note the Governing Jurisdiction","url":"https://github.com/deepseek-ai/DeepSeek-V3/blob/main/LICENSE-MODEL","hn":"https://news.ycombinator.com/item?id=42884909","source":"Hacker News","lang":"en","date":"2025-01-31","points":6,"comments":6},{"title":"DeepSeek-v3.2-Exp: Long-Context Efficiency with DeepSeek Sparse Attention [pdf]","url":"https://github.com/deepseek-ai/DeepSeek-V3.2-Exp/blob/main/DeepSeek_V3_2.pdf","hn":"https://news.ycombinator.com/item?id=45416796","source":"Hacker News","lang":"en","date":"2025-09-29","points":4,"comments":0},{"title":"Show HN: Tokiwi – An online tokenizer for any Hugging Face model","url":"https://www.tokiwi.dev/","hn":"https://news.ycombinator.com/item?id=43516781","source":"Hacker News","lang":"en","date":"2025-03-29","points":2,"comments":0},{"title":"DeepSeek Sparse Attention: Boosting Long-Context Efficiency [pdf]","url":"https://github.com/deepseek-ai/DeepSeek-V3.2-Exp/blob/main/DeepSeek_V3_2.pdf","hn":"https://news.ycombinator.com/item?id=45421845","source":"Hacker News","lang":"en","date":"2025-09-30","points":1,"comments":0},{"title":"DeepSeek v3: The Open-Source LLM Challenger Making Waves","url":"https://opentools.ai/news/deepseek-v3-the-open-source-llm-challenger-making-waves","source":"opentools.ai","lang":"en","date":"2026-09-12"},{"title":"DeepSeek v3.2: What's new and how does it compare to ChatGPT?","url":"https://mashable.com/article/deepseek-v3-2-models-released","source":"Mashable","lang":"en","date":"2025-12-03"},{"title":"DeepSeek-V3-0324 is now generally available in GitHub Models","url":"https://github.blog/changelog/2025-04-08-deepseek-v3-0324-is-now-generally-available-in-github-models/","source":"github","lang":"en","date":"2025-04-08"},{"title":"DeepSeek V3 Is Now Reportedly the Best Nonreasoning AI Model","url":"https://www.cnet.com/tech/services-and-software/deepseek-v3-is-now-reportedly-the-best-non-reasoning-ai-model/","source":"CNET","lang":"en","date":"2025-03-25"},{"title":"DeepSeek-V3 is now generally available in GitHub Models","url":"https://github.blog/changelog/2025-03-07-deepseek-v3-is-now-generally-available-in-github-models/","source":"github","lang":"en","date":"2025-03-07"},{"title":"DeepSeek Open-Sources DeepSeek-V3, a 671B Parameter Mixture of Experts LLM","url":"https://www.infoq.com/news/2025/01/deepseek-v3-llm/","source":"InfoQ","lang":"en","date":"2025-01-21"},{"title":"訓練成本不到 600 萬美元，中國 AI 公司開源 DeepSeek-V3 新模型","url":"https://technews.tw/2024/12/30/introducing-deepseek-v3/","source":"TechNews 科技新報","lang":"zh","date":"2024-12-30"},{"title":"Chinese AI firm releases DeepSeek V3, a new leader in open-source AI models","url":"https://www.thehindu.com/sci-tech/technology/chinese-ai-firm-releases-deepseek-v3-a-new-leader-in-open-source-ai-models/article69032888.ece","source":"thehindu","lang":"en","date":"2024-12-27"},{"title":"Chinese AI company DeepSeek releases 'DeepSeek-V3', an AI model comparable to GPT-4o, with an astounding 671 billion parameters","url":"https://gigazine.net/gsc_news/en/20241227-deepseek-v3","source":"GIGAZINE","lang":"en","date":"2024-12-27"},{"title":"DeepSeek’s new AI model appears to be one of the best ‘open’ challengers yet","url":"https://techcrunch.com/2024/12/26/deepseeks-new-ai-model-appears-to-be-one-of-the-best-open-challengers-yet/","source":"TechCrunch","lang":"en","date":"2024-12-26"},{"title":"DeepSeek-V3, ultra-large open-source AI, outperforms Llama and Qwen on launch","url":"https://venturebeat.com/ai/deepseek-v3-ultra-large-open-source-ai-outperforms-llama-and-qwen-on-launch","source":"VentureBeat","lang":"en","date":"2024-12-26"},{"title":"DeepSeek V3多項測試指標領先Llama，中國語言模型引發言論自由與話語權討論","url":"https://www.ithome.com.tw/news/166720","source":"iThome Online","lang":"zh","date":"2024-12-26"},{"title":"unsloth/DeepSeek-V3-bf16","url":"https://huggingface.co/unsloth/DeepSeek-V3-bf16","source":"huggingface","lang":"en","date":"2012-11-05"},{"title":"第一时间,寒武纪原生适配DeepSeek-V4！两大国产之光再次强强联合","url":"https://news.google.com/rss/articles/CBMiswJBVV95cUxPanZUblo5MlgzWGxXa2g2MjRuT25CUFpBS0t4UHp0cjlJNVVmOTM3VFQ5T3BZMkdXYzlKZmlmU0VOQWlscjFPMWdkT0p1RVVGQjVHSmZNWFdQa01BRC1CWjh0SDNQeVF0cFV5U2VaVmVEYl85cEszaWtVckRuUUNwdGZqb1NzZUdiMWhlM2ROellOZTZXamhueC0zMkptUnVuSFMteTYwdG9NLWprVjVnTHZObFpXWHVtM1ltQVpELXZFYTZYcGd4d2w0V2UzYm9yTUpobXUzLXVHdnJGZGRVRFZmVEFaTjBEWlF2a1pNYml5TDM0dWRERWlQN0g5QzgwWnhuLVQ2RmQ0V3huQkd5d1ZUMTJJQmFoRFZmRTlwV0VJLUNtVEl6ZE5TOFRrMEJKbGxj?oc=5","source":"新浪网","lang":"zh","date":"2026-04-24"},{"title":"DeepSeek连更GitHub，华尔街回想起被支配的恐惧：“DeepSeek第二时刻”要来了","url":"https://news.google.com/rss/articles/CBMiU0FVX3lxTE05WEhZVmhac2pldWdXM25saFBURUNWeEszZzNBLVBGNHZYUzM5dWpQQnQyLUVib3BVd215dGtfZFZ2Sk1QSmNKblo3OVVMYmtoT0xv?oc=5","source":"华尔街见闻","lang":"zh","date":"2026-02-24"},{"title":"DeepSeek AI Researchers Introduce Engram: A Conditional Memory Axis For Sparse LLMs","url":"https://news.google.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?oc=5","source":"MarkTechPost","lang":"en","date":"2026-01-14"},{"title":"刚刚，DeepSeek 突发梁文峰署名新论文：V4 新架构提前曝光？","url":"https://news.google.com/rss/articles/CBMiXkFVX3lxTFAwWjZSX21SdXBMQkJhS3lDdk1MYndzV0hLTVMyenNCZDZVc013dDBVWGpwVXlxOU0ycDBHZ3hKU3ZFVllfQk1DU0x1cDgzU3FqdEZEejhtclFSTEtSekE?oc=5","source":"InfoQ.cn","lang":"zh","date":"2026-01-13"},{"title":"DeepSeek新模型上线！引入DSA新稀疏注意力，还又狙了CUDA一枪","url":"https://news.google.com/rss/articles/CBMiSEFVX3lxTFBPZmp0dGdhbUlfOXI1RGdnUjZ6bXRoX1BxbnNCV0JjQjFYRjMtSWZKdmsxWHlSbDhyYU9Sc1VmTkJGWk9kbEppcA?oc=5","source":"智源社区","lang":"zh","date":"2025-10-01"},{"title":"刚刚，DeepSeek开源V3.2-Exp，公开新稀疏注意力机制DSA","url":"https://news.google.com/rss/articles/CBMiTkFVX3lxTE9SU0J5Z1RDeTNVWEp3SXJvMmJQWnhsVjBQREVvUC1UZ25WektPc0dPY3RFRGZXWGwtOGJPdmdwcFFlLVNhWDJDY1Z3cDl4dw?oc=5","source":"36氪","lang":"zh","date":"2025-09-29"},{"title":"Details about METR's preliminary evaluation of DeepSeek-R1","url":"https://news.google.com/rss/articles/CBMiXEFVX3lxTFBSNHo5b1lWME0xOE91dVRoLTdRMnRaWVNiYXpldHJWaU5pdUxwT3FMbWVRNEgxSXFtRHBDZW1zWmhfZXZibm9ocnNjbFN2cy05ZXBUQ2xrNmFJcGtE?oc=5","source":"METR","lang":"en","date":"2025-03-05"},{"title":"DeepSeek满血微调秘籍来了，全网首发打破低价内卷！解锁升级版全家桶","url":"https://news.google.com/rss/articles/CBMiVkFVX3lxTFA0aXFfd1Zpc0dpLTd5SmFWMWxLYXJBXzloZGpBNVRuaVFoeS10Rzc0MTRmRnI4U3FEOVVueVhFN1pqeGRNYnd1MGdPMGJLMWx5Y1ZLcG1B?oc=5","source":"腾讯新闻","lang":"zh","date":"2025-02-19"},{"title":"DeepSeek-V3: Technical Details","url":"https://news.google.com/rss/articles/CBMibEFVX3lxTE0xT2dHYWJXWS16MFFJSkpFX0FEWE0xbkVsdmJfaDVOOHc0UldYcGVDaEEtT2VWQkpqOUw0U3pFTEZJTXlRMXFlYjFyVlNuR0hZbmRyekFHYW4zaHhudVA2SXhZVDRkTWpQU1Q4cA?oc=5","source":"Gonzo ML","lang":"en","date":"2025-02-06"}],"stale":false,"snapshot_at":"2026-11-16T09:23:14Z","history_start":"2026-09-06"}