[{"_1":2,"_101":-5,"_102":-5},"loaderData",{"_3":4,"_6":7},"root",{"_5":-5},"user","routes/tag.$slug",{"_8":9,"_90":91,"_98":93,"_99":61,"_100":62},"posts",[10,70],{"_11":12,"_13":14,"_15":16,"_17":18,"_19":20,"_21":22,"_23":24,"_25":26,"_27":28,"_29":30,"_31":32,"_33":34,"_35":-5,"_36":37,"_38":39,"_49":50,"_63":64,"_65":66,"_67":68,"_69":32},"documentId","je939inq93g6ptsa3chtu1wm","title","Kai-Fu Lee's AI Model '01.AI API' Goes Online","slug","kai-fu-lees-ai-model-01-ai-api-goes-online","excerpt","On March 15th, it was reported that Kai-Fu Lee's 01.AI has recently launched its official API.","content","\n

On March 15th, it was reported that Kai-Fu Lee's 01.AI has recently launched its official API. Users can experience the Yi series AI large models, including yi-34b-chat-0205, yi-34b-chat-200k, and yi-vl-plus.

\n\n\n\n

yi-34b-chat-0205: A version deeply optimized based on the open-source edition, with a 30% increase in instruction compliance and significantly reduced model response latency. Suitable for scenarios such as chatting, Q&A, dialogue, collaboration, and translation.

\n\n\n\n

yi-34b-chat-200k: Supports processing approximately 200K characters of Chinese (equivalent to about one book of \"Harry Potter\") or English words. Suitable for understanding multiple document contents, massive data analysis and mining applications across different domains.

\n\n\n\n

yi-vl-plus: Supports high-resolution image input at 1024*1024 pixels with capabilities including image Q&A, chart comprehension, OCR, and visual reasoning. Suitable for analyzing complex charts and screenshots for tasks like information recognition, extraction, comprehension, reasoning etc.

\n\n\n\n

According to the introduction, the Yi-34B-Chat series API has a faster inference speed, which not only shortens processing time but also maintains excellent model performance. In addition, the optimized API interface significantly reduces model response latency, further enhancing user experience in terms of smoothness and response speed.

\n\n\n\n

Furthermore, the Yi Model API is fully compatible with OpenAI API, allowing users to smoothly migrate by making minor code modifications.

\n\n\n\n

\n","publishedDate","2024-03-16T00:31:35.000Z","featured",false,"viewCount",0,"readingTime",2,"status","draft","language","en","originalWordPressId",164131,"featuredImage","featuredImageUrl","https://assets.pandaily.com/uploads/2024/03/0.1AI.jpg","categories",[40],{"_11":41,"_42":43,"_15":44,"_45":46,"_47":48},"ou7tkn9cq6eba9jr68hqiwch","name","AI","ai","color","#000000","description","","tags",[51,56,59],{"_11":52,"_42":53,"_15":54,"_45":55,"_47":48},"j45fwu61bg7hx57fo10s37oq","China","china","#6366f1",{"_11":57,"_42":58,"_15":58,"_45":55,"_47":48},"kxf65hwq1gw6c7lke4uk58yj","api",{"_11":60,"_42":61,"_15":62,"_45":55,"_47":48},"w22l19333obyhsn1xwfyb80k","0.1AI","0-1ai","createdAt","2025-06-24T22:22:49.678Z","updatedAt","2025-06-24T23:50:38.095Z","publishedAt","2025-06-24T23:50:38.128Z","locale",{"_11":71,"_13":72,"_15":73,"_17":74,"_19":75,"_21":76,"_23":24,"_25":26,"_27":28,"_29":30,"_31":32,"_33":77,"_35":-5,"_36":37,"_38":78,"_49":80,"_63":87,"_65":88,"_67":89,"_69":32},"rs4bqgu53cmgo4lygxp7eebn","Kai-Fu Lee's AI Company \"01.AI\" Announces the Open Source of the Yi-9B Model","kai-fu-lees-ai-company-01-ai-announces-the-open-source-of-the-yi-9b-model","Kai-Fu Lee's AI company 01.AI announced the open source of the Yi-9B model, claiming it has the strongest code and mathematical capabilities in the same series.","\n

On March 6th, \"01.AI\" official WeChat account announced the open sourcing of Yi-9B model. The official referred to it as the \"top scorer in science\" among the Yi series models - Yi-9B is currently the most powerful model in terms of code and mathematical capabilities within the Yi series models, with actual parameters at 8.8 billion and a default context length of 4K tokens.

\n\n\n\n

The model is based on Yi-6B (trained with 3.1T tokens) and further trained with 0.8T tokens, data up to June 2023.

\n\n\n\n

According to the introduction, in terms of overall performance (Mean-All), Yi-9B's performance is \"the best among open-source models of similar size\", surpassing DeepSeek-Coder, DeepSeek-Math, Mistral-7B, SOLAR-10.7B and Gemma-7B.

\n\n\n\n

In terms of code capability (Mean-Code), Yi-9B's performance is second only to DeepSeek-Coder-7B, outperforming Yi-34B, SOLAR-10.7B, Mistral-7B and Gemma-7B.

\n\n\n\n

In terms of mathematical ability (Mean-Math), the Yi-9B is second only to the DeepSeek-Math-7B, outperforming the SOLAR-10.7B, Mistral-7B and Gemma-7B.

\n\n\n\n

In terms of common sense and reasoning ability (Mean-Text), the Yi-9B's performance is on par with the Mistral-7B, SOLAR-10.7B, and Gemma-7B.

\n\n\n\n

The official statement also mentioned that Yi-9B (BF 16) and its quantized version Yi-9B (Int8) can be easily deployed on consumer-grade graphics cards, with low usage costs and developer-friendly features.

\n\n\n\n

\"01.AI\" was founded by Kai-Fu Lee, Chairman and CEO of Sinovation Ventures. It has previously released two open-source large models, Yi-34B and Yi-6B, which are claimed to be completely open for academic research and simultaneously available for free commercial applications.

\n\n\n\n

\n","2024-03-07T21:37:38.000Z",163981,[79],{"_11":41,"_42":43,"_15":44,"_45":46,"_47":48},[81,82,86],{"_11":52,"_42":53,"_15":54,"_45":55,"_47":48},{"_11":83,"_42":84,"_15":85,"_45":55,"_47":48},"f1mlf7ca7fdcrqi6md5dx217","open source","open-source",{"_11":60,"_42":61,"_15":62,"_45":55,"_47":48},"2025-06-24T22:23:06.621Z","2025-06-24T23:45:32.593Z","2025-06-24T23:45:32.630Z","pagination",{"_92":93,"_94":95,"_96":93,"_97":28},"page",1,"pageSize",12,"pageCount","total","currentPage","tag","tagSlug","actionData","errors"]

Tag: #0.1AI

Kai-Fu Lee's AI Model '01.AI API' Goes Online

Kai-Fu Lee's AI Company "01.AI" Announces the Open Source of the Yi-9B Model