---
title: Moonshot says it stripped cyber-offense data from Kimi K3.1 Pro before the open-weight release, cutting its exploit pass@3 from 90.5% to 81%. That is still far above K3 | ctxsec
url: https://ctxsec.com/items/moonshot-kimi-k3-1-pro-cyber-data-removed-open-weights
description: ""
status: "confirmed (Moonshot K3.1 Pro tech blog); benchmark figures are Moonshot's own and not independently checked, last updated Oct 26, 2026, 8:05 AM ET"
published: 2026-10-23
updated: 2026-10-26T08:05:00-04:00
---

> 

[← AI](/ai) · [markdown](/items/moonshot-kimi-k3-1-pro-cyber-data-removed-open-weights.md)

confirmed (Moonshot K3.1 Pro tech blog); benchmark figures are Moonshot's own and not independently checked

2026-10-23 · updated

Oct 26, 2026, 8:05 AM ET

### Moonshot says it stripped cyber-offense data from Kimi K3.1 Pro before the open-weight release, cutting its exploit pass@3 from 90.5% to 81%. That is still far above K3

Moonshot released Kimi K3.1 Pro (3T-class, 1M context) in its apps and API on Oct 23 and says full weights will be published **by Nov 4**. Its tech blog says vulnerability-research skill rose steadily with post-training, traced it to cyber content in the mid-training corpus and cyber tasks in the RL mix, and removed both for the open release while keeping general coding.

- On Moonshot's in-house **xdev-bench-hard** (exploit real recent bugs in Linux, Node.js, JavaScriptCore, Exim and VirGL with mitigations on, to arbitrary code execution), the unfiltered checkpoint scored **90.5%** pass@3 and the released model **81.0%**. Per TNW, K3 scored 33.3%.
- Moonshot says its policy is not to release open weights that extend the absolute frontier of cyber capability, and that K3.1 Pro stays behind frontier proprietary models that have been public for at least three months. It also says it intends to "progressively raise" the cyber capability available in open weights.

Defender takeaway: once the weights ship, a model that by its maker's own numbers exploits most of a hard real-bug benchmark will run with no usage monitoring. Data filtering reduced the capability but didn't remove it. Plan for faster exploit development against newly disclosed bugs, not for a model that can't do it.

Sources: [Moonshot AI: Kimi K3.1 Pro tech blog (cybersecurity section)](https://www.kimi.ai/blog/kimi-k3-1-pro) · [The Next Web: Moonshot releases Kimi K3.1 Pro and says it cut its hacking skills (Oct 23)](https://thenextweb.com/news/moonshot-kimi-k3-1-pro-open-weight-cyber)