Inference
EN
Xiaomi

MiMo-V2-Flash

MiMo-V2-Flash is a reasoning model from Xiaomi. 13 benchmarks count toward its index, in 6 categories.

  • Open weights
  • Reasoning
  • Input: text
  • Released 2026-03
RankedEnough results for an index.
IndexAll categories, out of 100.43.4±6.0Median 51.0
CoverageIndex weight that has results.85%
SpeedOutput tokens per second.129/sMedian 67/s
First tokenSeconds to the first output token.2.1 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.—Median $0.60
Output / 1MUS dollars per 1M output tokens.—Median $2.55
ContextMaximum tokens in one request.256KMedian 262K
EloLMArena rating and rank.—Median 1422

Run locally

309.8B parameters
16-bit
651.6 GB
8-bit
346.6 GB
4-bit
198.2 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy needs more than 128 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.2 counted48.5141 / 218
CodingCode writing and repair.2 counted53.0119 / 209
ReasoningLogic problems and puzzles.3 counted40.4158 / 227
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.4 counted38.0209 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted26.3—
MathMath problems.1 counted50.1—

Results

13 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
American Invitational Mathematics Examination 2025Math94.1%10 / 53
Massive Multitask Language Understanding ProfessionalKnowledge84.9%22 / 75
Graduate-Level Google-Proof Q&AKnowledge83.7%58 / 141
τ²-Bench Tool-Agent-User EvaluationAgentic83.9%62 / 149
Software Engineering Benchmark VerifiedCoding73.4%53 / 114
LiveCodeBench v6Coding80.6%28 / 59
LongBench v2Reasoning60.6%14 / 25
Artificial Analysis Humanity's Last ExamKnowledge8.6%152 / 222
Critical Physics TasksReasoning0.0%150 / 214
GDPval-AA normalizedAgentic6.4%94 / 133
Artificial Analysis IFBenchInstruction39.9%104 / 146
Artificial Analysis GPQA DiamondKnowledge65.6%142 / 199
Artificial Analysis Long Context ReasoningReasoning36.3%173 / 217
GPQA Diamond · Vals AIKnowledge59.3% ±2.5110 / 135
Artificial Analysis Omniscience AccuracyKnowledge15.6%149 / 164

Sources

BenchLMCC BY-NC 4.0Vals AINo licence stated

History

5 releases
Index43.5 → 43.4
23 Oct 202612 Nov 2026

Questions

How good is MiMo-V2-Flash?
Its index is 43.4 out of 100. That is place 189 of 283 ranked models.
What is MiMo-V2-Flash best at?
Coding: 53.0 out of 100, place 119 of 209.
How fast is MiMo-V2-Flash?
129 output tokens per second, 2.1 s to the first token.
What is the context window of MiMo-V2-Flash?
256K tokens.
Is MiMo-V2-Flash open source?
Yes. Its weights are published.
When was MiMo-V2-Flash released?
March 2026.

Badge

MiMo-V2-Flash on Inference360: place 189 of 283, index 43.4
HTML
<a href="https://inference360.ai/models/mimo-v2-flash"><img src="https://inference360.ai/badges/mimo-v2-flash.svg" alt="MiMo-V2-Flash on Inference360: place 189 of 283, index 43.4" height="20"></a>
Markdown
[![MiMo-V2-Flash on Inference360: place 189 of 283, index 43.4](https://inference360.ai/badges/mimo-v2-flash.svg)](https://inference360.ai/models/mimo-v2-flash)

Compare with

Trinity-Large-Thinking43.5GPT-OSS 120B43.4MiMo-V2-Pro51.8

More from Xiaomi

All models
MiMo-V2-Pro51.8MiMo-V2-Omni47.2MiMo-V2.5-Pro56.3MiMo-V2.554.8MiMo V2.6 Pro UltraSpeedUnranked