Kimi K3 Goes Open-Source, Hits Leaderboards
roll0ver · reddit · 2026-07-20
The post states that Moonshot released Kimi K3:
- Approximately 2.8 trillion parameters
- Fully open-source / open-weight
- On Artificial Analysis's independent leaderboard, outperforms Anthropic's Opus 4.8 on some benchmarks
- Still trails behind Claude Fable 5 and GPT-5.6 overall
The author further notes that such results have impacted market sentiment: several Chinese AI companies' stocks dropped, and Nvidia briefly lost its top market cap spot intraday; meanwhile, Moonshot plans to push for an IPO in about 6 months with a target valuation of over $30 billion.
Related event: Kimi K3 Released Open-Source: 2.8T Parameters Shakes the Industry(3 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