AI在线 AI在线

Perplexity CEO Praises Kimi K2 Model and May Conduct Further Training

Recently, Aravind Srinivas, CEO of the US artificial intelligence startup Perplexity, posted an important message on social media. He stated that the company highly praised the Kimi K2 model launched by "Moonshot," and expressed the possibility of further training based on K2 in the future.

Recently, Aravind Srinivas, CEO of the US artificial intelligence startup Perplexity, posted an important message on social media. He stated that the company highly praised the Kimi K2 model launched by "Moonshot," and expressed the possibility of further training based on K2 in the future. This news has attracted widespread attention.

Perplexity CEO Praises Kimi K2 Model and May Conduct Further Training

Image source note: The image was generated by AI, and the image licensing service provider is Midjourney

Kimi K2 is the first trillion-parameter open-source model released by "Moonshot," focusing on tasks such as code generation and general intelligent agents (Agent). According to the official introduction, Kimi K2 has achieved excellent results in multiple performance benchmark tests, including SWE Bench Verified, Tau2, and AceBench. These test results indicate that Kimi K2 performs well in code, intelligent agents, and mathematical reasoning, demonstrating leading capabilities.

The architecture of Kimi K2 adopts a mixture of experts (MoE) design, with a total parameter count of an impressive 1 trillion, and an activated parameter count of 32 billion. Such a parameter scale enables it to better handle challenges in various complex tasks.

It is reported that Perplexity had previously used the DeepSeek R1 model for training, which laid the foundation for its further exploration in the field of artificial intelligence. Aravind's positive evaluation of Kimi K2 indicates that Perplexity is actively seeking new technology and model collaboration opportunities to promote the development of its products and services.

The release of Kimi K2 not only enhances "Moonshot"'s competitiveness in the open-source model field but also provides new possibilities for Perplexity's future development. With the rapid development of artificial intelligence technology, companies that can maintain a leading position in this field will bring more innovative solutions to the market.

Aravind's statement shows Perplexity's confidence in the Kimi K2 model and its expectations and strategies for future developments in artificial intelligence technology.

Key Points:

🌟 Kimi K2 is the first trillion-parameter open-source model released by "Moonshot," excelling in code and intelligent agent tasks.  

🚀 Perplexity CEO praised the Kimi K2 model and plans for subsequent training.  

📈 Kimi K2 has achieved outstanding results in multiple benchmark tests, demonstrating strong capabilities.

相关资讯

Major Release! Moonshot Introduces Open-Source Large Model Kimi K2 with Trillion Parameters

Chinese AI company Moonshot AI recently launched its latest open-source large language model — Kimi K2. This model adopts a mixture of experts architecture, with a total parameter count of an impressive 1 trillion, of which 32 billion parameters are activated, demonstrating excellent performance and becoming a focus of attention in the industry.Kimi K2 not only impresses with its scale of parameters, but also boasts powerful agent capabilities.
7/14/2025 2:02:00 PM
AI在线

The Ministry of Industry and Information Technology will release the 'International Artificial Intelligence Open Source Cooperation Initiative' at the 2025 World Artificial Intelligence Conference

The 2025 World Artificial Intelligence Conference will be held in Shanghai from July 26 to 28, 2025. To this end, the Shanghai Municipal Government's Information Office held a press conference on July 10 to introduce the progress of the conference preparations.
7/14/2025 2:01:51 PM
AI在线

Moon's Dark Side Kimi K2 API is Slow, Being Optimized Intensively!

Recently, the official Weibo account of Moonshot AI responded to the issue of slow speed for Kimi K2 API, attracting a lot of user attention. According to Moonshot AI, the main reasons for the slow speed are a sharp increase in traffic and the large model size.
7/16/2025 4:11:50 PM
AI在线
  • 1