Moonshot AI releases Kimi K3.1 Pro with 1M-token context and open weights by Nov 4

Moonshot AI reports that K3.1 Pro scored 65.7% on Terminal-Bench 4.0, ahead of GPT-6 Astra at 58.2% and narrowly behind Claude Opus 5.5 at 66.4%. On Kimi Code Bench 3.0 Long-horizon, it scored 61.2% versus K3's 55.95%. On BenchCAD for 3D reconstruction, it reached 92.7% with tools, approaching Claude Opus 5.5 at 96.2% and GPT-6 Astra at 95.9%. The company also reports a state-of-the-art 64.4 on OfficeQA Pro v2 for fiscal data analysis, and a reduction in claim-level hallucination rate from 23.4% for K3 to 8.3% on its internal Kimi Hallu Bench Hard. A 3D Make feature is live on kimi.ai and the Kimi app for previewing, editing, and downloading 3D models for printing.

For the open-weight release, Moonshot AI said it identified and removed cyber-related data from training while retaining general-purpose coding tasks. On xdev-bench-hard, a checkpoint without risk-reduction measures solved 90.5% of problems at pass@3, while the final K3.1 Pro reaches 81.0%. The company states that K3.1 Pro's cyber capability remains behind frontier proprietary models available for at least three months. Full model weights for Kimi K3.1 Pro will be released by November 4, 2026.