01.AI Releases Yi Large Model API and Launches Public Beta: Supports 200K Context
-
01.AI has released the Yi large model API and launched a public beta. This beta test provides two models: Yi-34B-Chat (0205) and Yi-34B-Chat-200K.
Among them, Yi-34B-Chat-200K supports ultra-long context processing, achieving 200K context support, capable of handling approximately 200,000 to 300,000 Chinese and English characters. It is suitable for scenarios such as multi-document content understanding, massive data analysis and mining, and cross-domain knowledge integration. Additionally, Yi-34B-Chat-200K performs excellently in handling complex roles and relationships, such as summarizing intricate characters and relationships in lengthy and complex literary works like Wuthering Heights.
Yi-34B-Chat (0205) has undergone deep optimization, significantly improving its instruction-following and creative content generation capabilities. It better understands user needs, adheres to multi-constraint instructions, and stably generates content in specified formats. For example, when outputting JSON-formatted book lists or determining text sentiment tendencies, Yi-34B-Chat (0205) performs more stably and accurately compared to previous versions. The Yi Large Model API has also been optimized in terms of inference speed, offering faster processing times while maintaining outstanding model performance. The optimized API interface reduces response latency, improving the smoothness and responsiveness of the user experience. Furthermore, the Yi Large Model API is fully compatible with the OpenAI API, enabling seamless migration to experience the extraordinary capabilities of the Yi model.
Application Address: https://01ai.feishu.cn/share/base/form/shrcn6cLrvS0jQKMgPMwmMdd06o