Inference
EN
Xiaomi

MiMo-V2.5

MiMo-V2.5 is a reasoning model from Xiaomi. 14 benchmarks count toward its index, in 5 categories.

  • Open weights
  • Reasoning
  • Input: text · audio · image · video
  • Released 2026-04
RankedEnough results for an index.
IndexAll categories, out of 100.54.8±6.8Median 51.0
CoverageIndex weight that has results.75%
SpeedOutput tokens per second.4/sMedian 67/s
First tokenSeconds to the first output token.78.5 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.40Median $0.60
Output / 1MUS dollars per 1M output tokens.$2.00Median $2.55
ContextMaximum tokens in one request.1MMedian 262K
EloLMArena rating and rank.1428±4 #102Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
xiaomi/mimo-v2.5
Max outputMaximum output tokens per request.
66K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.08
WeightsLink to the published weights.
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
4 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
Deep InfraDeprecatedHost$0.14$0.28262K
XiaomiDeprecatedMaker$0.14$0.281.05M
Novita AIHost$0.168$0.3361.05M
OpenRouterDeprecatedRouter$0.40$2.001M

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Run locally

310.76B parameters
16-bit
653.6 GB
8-bit
347.7 GB
4-bit
198.8 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.4 counted56.592 / 218
CodingCode writing and repair.5 counted55.591 / 209
ReasoningLogic problems and puzzles.None counted——
MultimodalTasks with images and text.2 counted57.248 / 84
KnowledgeFacts and expert knowledge.2 counted50.5132 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.1 counted52.7—

Results

14 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
Terminal-Bench 2.0 · BenchLMAgentic65.8%11 / 41
LiveCodeBenchCoding81.5% ±1.160 / 142
GPQA Diamond · Vals AIKnowledge81.6% ±2.161 / 135
MMMU ProMultimodal80.0% ±1.045 / 92
Gert Labs Composite Game BenchmarkAgentic46.9%26 / 52
Massive Multi-discipline Multimodal Understanding ProMultimodal77.9%28 / 53
Terminal-Bench 2.1 · Vals AIAgentic60.7% ±5.240 / 75
MMLU ProKnowledge82.9% ±0.475 / 135
Claw-EvalAgentic62.3%24 / 42
Vibe Code Bench v1.1Coding42.2% ±4.675 / 123
CharXiv ReasoningMultimodal81.0%30 / 49
SWE-bench ProCoding56.1%51 / 83
SWE-benchCoding71.0% ±2.060 / 87
Code MigrationCoding14.2% ±3.757 / 77
ProofBench v1.1Math16.0% ±3.746 / 54
TaxEval v2Finance71.8% ±0.972 / 143
CorpFin v2Finance59.9% ±1.071 / 131
EMBFinance55.1% ±3.150 / 75
LegalBenchLegal78.9% ±0.5101 / 149
MortgageTaxFinance59.3% ±1.067 / 97
MedScribeHealthcare72.2% ±1.990 / 109
Finance Agent (v2)Finance36.7% ±0.167 / 79
MedCodeHealthcare31.9% ±2.091 / 106
Tax Agent BenchFinance7.8% ±1.871 / 82
Legal Research BenchLegal9.1% ±2.082 / 90
Video-MME with subtitleMultimodal87.7%—
MM-ClawBenchAgentic23.8%—
ResearchClawBenchAgentic16.9%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseVals AINo licence statedmodels.devMIT

History

5 releases
Index55.7 → 54.8
23 Oct 202612 Nov 2026
Output / 1M$0.28 → $2.00
23 Oct 202612 Nov 2026

Questions

How much does MiMo-V2.5 cost?
$0.40 per 1M input tokens, $2.00 per 1M output tokens.
How good is MiMo-V2.5?
Its index is 54.8 out of 100. That is place 105 of 283 ranked models.
What is MiMo-V2.5 best at?
Agentic: 56.5 out of 100, place 92 of 218.
How fast is MiMo-V2.5?
4 output tokens per second, 78.5 s to the first token.
What is the context window of MiMo-V2.5?
1M tokens.
Is MiMo-V2.5 open source?
Yes. Its weights are published.
When was MiMo-V2.5 released?
April 2026.

Badge

MiMo-V2.5 on Inference360: place 105 of 283, index 54.8
HTML
<a href="https://inference360.ai/models/mimo-v2-5"><img src="https://inference360.ai/badges/mimo-v2-5.svg" alt="MiMo-V2.5 on Inference360: place 105 of 283, index 54.8" height="20"></a>
Markdown
[![MiMo-V2.5 on Inference360: place 105 of 283, index 54.8](https://inference360.ai/badges/mimo-v2-5.svg)](https://inference360.ai/models/mimo-v2-5)

Compare with

Inkling54.8Qwen3.6 Plus54.6GPT-555.0MiMo-V2.5-Pro56.3dots3-note Preview63.4

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeGLM-5.3-Flash63.1 · $0.078/1M

More from Xiaomi

All models
MiMo-V2.5-Pro56.3MiMo-V2-Pro51.8MiMo-V2-Omni47.2MiMo-V2-Flash43.4MiMo V2.6 Pro UltraSpeedUnranked