Benchmark Almanac

Xiaomi MiMo

We track 5 benchmarks that Xiaomi MiMo made or helped make.

3 in use 1 with public code 4 with a leaderboard here

Benchmarks

Name Kind Released Best score Status
MiMo Visual Coding Models build websites from requests and turn reference images into matching front-end code. Tasks include WebDev and Image2Code. Web and app front ends Sep 2026 82.2% In use
MiMo Code Bench Coding agents work through a varied set of software tasks built in house by Xiaomi. The report gives no task count or grading details. Fixing real repositories Sep 2026 68.6% In use
MiMo Cyber Bench Agents do in-house cybersecurity tasks that add to public vulnerability reproduction and exploitation benchmarks. The report gives no task count or grading details. Finding and exploiting real bugs Sep 2026 81.7% In use
MiMo Coding Bench Models do varied coding tasks inside agent harnesses such as Claude Code. Tasks cover repo understanding, project building, code review, artifact generation, planning and SWE. Fixing real repositories Apr 2026 77.1% Fading
SpeechMMLU Speech-language models answer MMLU-derived questions from text or synthesized speech in text-to-text, speech-to-text, text-to-speech and speech-to-speech settings. Audio understanding and voice Sep 2025 Not in cards