Moonshot launches Kimi K3, a 2.8T-parameter open-weights vision-language model
The Batch (Andrew Ng) · rss · 2026-07-24
Moonshot launched Kimi K3, a 2.8T-parameter vision-language model, and said weights will follow on July 27, which would make it the largest open-weights model yet.
The model accepts text, images, and video with up to 1M-token inputs and outputs, offers adjustable reasoning and tool use, and is already available via API and apps. It ranks first among open models on Artificial Analysis’ Intelligence Index, leads Arena.ai’s Code Arena WebDev leaderboard, and is priced at $3/$0.30/$15 per million input/cached/output tokens.
More from Models
- Daily AI brief: GPT-Live-1 in API, OpenAI pauses $200 Pro signups amid Astra demand — koltregaskes · 2026-09-11
- 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