Kimi K3: Open Frontier Intelligence
Kimi Team, Tongtong Bai, Yifan Bai, Yiping Bao et al. · arXiv · 2026
Abstract
Moonshot's technical report on Kimi K3, an open-weight mixture-of-experts model with 2.8 trillion total parameters but only 104 billion active per token, a million-token context, and vision built in rather than bolted on. The report covers the attention and routing changes behind roughly 2.5x better scaling efficiency than its predecessor, the reinforcement-learning recipe used for agentic and coding behaviour, and the infrastructure a model this size demands.
Why it matters
The clearest current look at how an open-weight lab builds a frontier-scale model — sparsity, long context, and agentic RL documented in one place. Read it when you want the engineering behind the leaderboard numbers rather than the numbers themselves.
https://arxiv.org/abs/2607.24653