Dean Ball says Kimi K3 is near the best public models for agentic coding, but token-hungry
FinanceYF5 · x · 2026-07-21
In a follow-up to Dean Ball’s Kimi thread, the author says the model is not just “distilled away” into seeming good: in agentic coding, K3 is close to the best public models available in early 2026.
The main caveat from his own use is that the model appears to be very token-hungry, so he is not convinced it is truly cheap to run in practice.
Because the claim is specifically about performance in coding-agent workflows, the post is more about how the model behaves in use than about the model announcement itself.
Related event: Dean Ball’s Kimi Assessment Fuels Debate on Open-Weight AI(6 posts)→
More from coding & agent
- Inspired by OpenAI's 10,000-agent run, dev open-sources a crowdsourced agent problem-solving platform — Benjaminsen · 2026-09-11
- Lucid: open-source Mac app keeps your laptop awake only while AI agents run — Pitiful_Hedgehog_600 · 2026-09-11
- banteg's snail project crowdsources AI agents to finish matching Snail Mail's 20 remaining functions — banteg · 2026-09-11
- Alex Townsend posts 200 open problems in numerical linear algebra for humans and AI agents — IgorCarron · 2026-09-11
- Kimi K2.8 Preview rolls out: near-K3 coding performance, 1M context for all tiers — teortaxesTex · 2026-09-11
- Looking for a classifier of software engineering task shapes to pick models per task — StewartalsopIII · 2026-09-11