Rankings / xiaomi
Mimo V2 Flash
released 2025-12-16
BenchAtlas Index
as of 2026-07-14
thinking
rank #143
9 families · 5 categories · high
no reasoning
rank #262
9 families · 5 categories · high
Benchmark evidence
25 results
agentic coding
| τ²-Bench | thinking | 95.0 % | independent | Artificial Analysis |
| τ²-Bench | no reasoning | 83.9 % | independent | Artificial Analysis |
| τ²-Bench subset=banking | no reasoning | 3.1 % | independent | Artificial Analysis |
coding
| LiveCodeBench v6 implementation=artificial-analysis | thinking | 86.8 % | independent | Artificial Analysis |
| LiveCodeBench v6 implementation=artificial-analysis | no reasoning | 40.2 % | independent | Artificial Analysis |
| SciCode | thinking | 39.4 % | independent | Artificial Analysis |
| SciCode | no reasoning | 25.9 % | independent | Artificial Analysis |
external indices
| AA Coding Index | no reasoning | 49.8 points | independent | Artificial Analysis |
| Intelligence Index v4.1 | thinking | 31.2 points | independent | Artificial Analysis |
| Intelligence Index v4.1 | no reasoning | 24.7 points | independent | Artificial Analysis |
| AA Math Index | thinking | 96.3 points | independent | Artificial Analysis |
| AA Math Index | no reasoning | 67.7 points | independent | Artificial Analysis |
knowledge science
| GPQA Diamond implementation=artificial-analysis | thinking | 84.6 % | independent | Artificial Analysis |
| GPQA Diamond implementation=artificial-analysis | no reasoning | 65.6 % | independent | Artificial Analysis |
| GPQA Diamond implementation=vals-ai | 59.3 % | independent | Vals AI | |
| Humanity's Last Exam implementation=artificial-analysis | thinking | 21.1 % | independent | Artificial Analysis |
| Humanity's Last Exam implementation=artificial-analysis | no reasoning | 8.0 % | independent | Artificial Analysis |
| MMLU-Pro implementation=artificial-analysis | thinking | 84.3 % | independent | Artificial Analysis |
| MMLU-Pro implementation=artificial-analysis | no reasoning | 74.4 % | independent | Artificial Analysis |
long context instruction
| AA-LCR | thinking | 63.0 % | independent | Artificial Analysis |
| AA-LCR | no reasoning | 31.3 % | independent | Artificial Analysis |
| IFBench | thinking | 64.2 % | independent | Artificial Analysis |
| IFBench | no reasoning | 39.9 % | independent | Artificial Analysis |
reasoning math
| AIME year=2025 · implementation=artificial-analysis | thinking | 96.3 % | independent | Artificial Analysis |
| AIME year=2025 · implementation=artificial-analysis | no reasoning | 67.7 % | independent | Artificial Analysis |
Agent + model results
systems, not bare-model scores
| agent + model Artificial Analysis harness + Mimo V2 Flash | Terminal-Bench 2.1 | 61.8 % | independent | Artificial Analysis |
| agent + model Artificial Analysis harness + Mimo V2 Flash | Terminal-Bench Hard | 28.0 % | independent | Artificial Analysis |
| agent + model Artificial Analysis harness + Mimo V2 Flash | Terminal-Bench Hard | 25.8 % | independent | Artificial Analysis |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.