Skip to content
EventPast event

Researcher says Xiaomi's open-source model ranks third in cyberattack capability evaluation

1 reports1 reporting sourcesLatest report

Start here

AI overview

On October 29, 2026, researcher Sergei Smirnov published an evaluation of the cyber capabilities of Xiaomi's open-source model MiMo-V2.6-Pro on LessWrong. He said the model ranked 3rd among 9 models on the 41 CVEs in ExploitBench, behind only Mythos (preview) and GPT-5.5. In the tiered results, MiMo reached vulnerability triggering (T4) or above on 33 CVEs, limited memory control within the V8 sandbox (T3) on 16, arbitrary read/write outside the sandbox (T2) on only 1, and never achieved code execution.

Generated by AI from reports · Updated 14 days ago

Latest update
Tracking the cyberattack capabilities of cheap open-source models

Related research

Papers whose work the reports mention

Progress timeline

Oct 29
  1. X@seriozh1· Discussion
    Xiaomi's recently released MiMo-V2.6-Pro is extremely cheap while performing well on cyber-related tasks. We evaluated…(opens the original post in a new tab)
  2. LessWrongSelected
    Tracking the cyberattack capabilities of cheap open-source models

    Researcher Sergei Smirnov published an evaluation on LessWrong of the cyber capabilities of Xiaomi's open-source model MiMo-V2.6-Pro, stating that it ranked 3rd among 9 models on 41 CVEs in ExploitBench, behind only Mythos (preview) and GPT-5.5. In the tiered results, MiMo reached vulnerability triggering (T4) or above on 33 CVEs, reached restricted memory control within the V8 sandbox (T3) on 16, achieved arbitrary read/write outside the sandbox (T2) on only 1, and never reached code execution (T1). The author used successful solutions from GLM-5.3-Prime to extract prompts as an oracle tool, enabling MiMo to reach T2 in 8 of 9 runs, compared with 1 run when operating alone. On an isolated unpatched PHP server, MiMo found PHP2Shell (CVE-2026-109618) online and read /flag.txt within about 13 steps, 4 minutes, and about 20 cents of inference cost.

Related discussion

Total 1 items