Moonshot AI Releases 2.8T Parameter Open-Source Model Kimi K3
AI寒武纪 · wechat · 2026-07-16
Moonshot AI has officially launched Kimi K3, a new open-source model with 2.8 trillion parameters, making it the largest publicly available open-source model to date. Built on a hybrid linear attention mechanism, it natively supports visual understanding and features a 1 million token context window, primarily targeting software engineering, knowledge work, and deep reasoning.
Benchmark Performance: Officially ranked third in overall intelligence, just behind Claude Fable5 and GPT-5.6 Sol. It scored 1687 on the GDPval-AAv2 benchmark for real-world work and demonstrated exceptional capabilities in long-horizon information retrieval tasks.
Key Features: K3 shows significant improvements in coding and agentic tasks. It can autonomously understand codebases, coordinate tools, and adjust strategies with minimal human intervention. It is particularly adept at engineering tasks combining visual understanding, such as frontend or game development. Model weights and a detailed technical report will be released soon.
Related event: Kimi K3 Debuts Strong, Narrowing the Open-Weight Gap(184 posts)→
More from Models
- Same Echo Maze prompt, three frontier models: all passed visually but shipped the same hidden bug — eyishazyer · 2026-09-11
- Benchmark scores drop from 89% to 19% on new evals — how benchmaxxing breaks leaderboard trust — airesearch12 · 2026-09-11
- ChatGPT tells user their question is too hard and to 'accept dumber answers' — phido3000 · 2026-09-11
- Developer Building a Unified Leaderboard of All Model Benchmark Scores — airesearch12 · 2026-09-11
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11
- GPT-5.6 writes well but is instantly forgettable, user complains — BasedRaddka · 2026-09-11