‹ PublicAI Index
The LLM benchmark aggregator.
Mimo V2.6 Flash
Xiaomi · 311B · open weights
Strongest in Cybersecurity (#1 of 9), weakest in Safety (#282 of 337). Above par in 11 of 16 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5 and Mimo V2.6 Pro and ahead of Gemini 3.6 Flash and Kimi K2.5.
§ 1 · Profile
What it is good at
Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.
Professional56.4−5.2#24/1681/1
Cybersecurity59.7leads#1/91/1
Legal57.9−9.2#22/1511/1
Finance56.5−6.4#30/1511/1
Coding57−12.1#24/1651/5
Agentic coding58.1−10.6#21/1571/4
Human preference62.2−5.3#58/3421/1
Human preference62.2−5.3#58/3421/1
Reasoning51.8−15.1#77/1781/4
Mathematics52.7−13#68/1401/2
Safety45.5−16#282/3371/3
Toxicity avoidance54−5#130/2721/1
Secure code48.4−18.1#174/2741/1
Harm refusal46.3−15.3#222/3001/2
Jailbreak resistance34−30.9#240/2721/1
Fairness42.6−27.4#252/3001/2
§ 2 · Head to head
What it beats, and what beats it
The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Mimo V2.6 Flash, left for the other.
§ 3 · Sources
Where the numbers come from
3 publications, 14 figures. Every one links to the page it was read from.
LMArena Text 1454
Vals · Legal Research Bench 37.98%Vals · Harvey Legal Agent Benchmark 11.25%Vals · Finance Agent 56.28%Vals · CyberBench 75.36%Vals · Vibe Code Bench 78.96% (OpenHands)Vals · Code Migration 40.93%Vals · ProofBench 63%
Enkrypt · Jailbreak risk 26.3%Enkrypt · Harmful content risk 2.2%Enkrypt · CBRN risk 51.7%Enkrypt · Toxicity risk 3.5%Enkrypt · Bias risk 89.4%Enkrypt · Insecure code risk 36.9%
Badge
[](https://publicai.io/model-index/m/mimo-v2-6-flash)