Today on the AI Daily Brief, NLW explores the rise of Kimi K2 Thinking, a new open-source model from China that’s outperforming GPT-5 and Claude 4.5 Sonnet on agentic benchmarks—and doing it at a fraction of the cost. We’ll look at how this shift is changing the balance of power between closed and open models, why Silicon Valley startups are already adopting Chinese systems, and what it means for the next phase of the AI race. Plus: Meta’s new speech model, DeepSeek’s dire job-market warning, and CoreWeave’s data-center delays.

Brought to you by:

Rovo - Unleash the potential of your team with AI-powered Search, Chat and Agents - ⁠⁠⁠⁠⁠⁠https://rovo.com/⁠⁠⁠⁠⁠⁠

AssemblyAI - The best way to build Voice AI apps - ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://www.assemblyai.com/brief⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Robots & Pencils - Cloud-native AI solutions that power results ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://robotsandpencils.com/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

The AI Daily Brief helps you understand the most important news and discussions in AI. Subscribe to the podcast version of The AI Daily Brief wherever you listen: https://pod.link/1680633614

Interested in sponsoring the show? sponsors@aidailybrief.ai

The Open Source AI Model Beating GPT-5 on Agents

The AI Daily Brief: Artificial Intelligence News and Analysis