Kimi K3: Strong Open Model, but Don’t Overhype It
Don't Worry About the Vase (Zvi) · rss · 2026-07-20
A long analysis of Kimi K3 argues that it is a very strong open model, but should not be overhyped.
Main points:
- It looks excellent on benchmarks, and if the planned weights release happens, it could be the strongest open model by raw capability.
- The author thinks it is still several months behind the closed frontier in aggregate practical capability.
- It is large and slow (2.8T), and much of the enthusiasm comes from that “big model smell.”
- Benchmarks likely overstate real-world performance because they are run at maximum effort.
- It may be especially strong for agentic coding and 3D, but will not replace cheaper smaller open models or top closed models in many workflows.
The piece also revisits the broader “DeepSeek moment” narrative and warns against drawing sweeping conclusions from one Chinese model release.
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
- Claude is no longer available for minors as Anthropic rolls out age assurance — Muhammad523 · 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