Skip to main content

Model profile

MiMo-V2.5

XiaomiCurrentReleased Apr 22, 2026
Data verified
Overall Score
58.62Public #62 of 200
Arena Elo
1433
Eligible category ranks
3of 8
Price (1M tokens)
Not listedAPI pricing
Speed
Not listed
Context
1M

Evidence coverage

11 of 323 tracked benchmarks are published. 10 are verified and 1 provisional. 3 of 8 categories are measured.

Updated July 21, 2026Methodology
Published / tracked
11 / 323
Verified
10
Provisional
1
Categories with evidence
3 / 8

Evidence by category

  • Agentic5 benchmarks
    Verified
  • Coding2 benchmarks
    Verified
  • Reasoning0 benchmarks
    Not measured
  • Knowledge0 benchmarks
    Not measured
  • Math0 benchmarks
    Not measured
  • Multilingual0 benchmarks
    Not measured
  • Multimodal4 benchmarks
    Mixed evidence
  • Inst. Following0 benchmarks
    Not measured
ProprietaryReasoning
Confidence:
Medium
base

MiMo-V2.5 ranks #62 out of 200 models on the public leaderboard with an overall score of 58.62/100. It does not yet have enough sourced coverage for BenchLM's verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.

MiMo-V2.5 is a proprietary model with a 1M token context window. It uses explicit chain-of-thought reasoning, which typically improves performance on math and complex reasoning tasks at the cost of higher latency and token usage.

MiMo-V2.5 sits inside the MiMo-V2.5 family alongside MiMo-V2.5-Pro. BenchLM links it directly to MiMo-V2-Omni as the earlier related model in that lineage. This profile currently has 11 of 323 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.

Its strongest category is Multimodal & Grounded (#19), while its weakest is Coding (#60). This performance profile makes it particularly strong for screenshots, documents, charts, and grounded multimodal workflows.

Peer position

Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.

Range 58.1559.0

  1. GPT-5.2 Instant
    OpenAI
    #5959.0
    GPT-5.2 Instant is #59 with a score of 59.0.
    Compare
  2. GPT-5.3 Instant
    OpenAI
    #6058.9
    GPT-5.3 Instant is #60 with a score of 58.9.
    Compare
  3. DeepSeek V4 Flash
    DeepSeek
    #6158.88
    DeepSeek V4 Flash is #61 with a score of 58.88.
    Compare
  4. MiMo-V2.5Current model
    Xiaomi
    #6258.62
    MiMo-V2.5 is #62 with a score of 58.62.
  5. GPT-5 (high)
    OpenAI
    #6358.61
    GPT-5 (high) is #63 with a score of 58.61.
    Compare
  6. GPT-5.2
    OpenAI
    #6458.43
    GPT-5.2 is #64 with a score of 58.43.
    Compare
  7. DeepSeek V3.2 (Thinking)
    DeepSeek
    #6558.15
    DeepSeek V3.2 (Thinking) is #65 with a score of 58.15.
    Compare

Category percentile

More

Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.

  1. Multimodal36%
    Eligible cohort rank #19 of 29Category score 63.3
  2. Agentic77%
    Eligible cohort rank #28 of 119Category score 53.8
  3. Coding51%
    Eligible cohort rank #60 of 122Category score 50.6

Category evidence

Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticRank #28 of 119Percentile 77thWeight 22%5 benchmarksVerified53.8
CodingRank #60 of 122Percentile 51stWeight 20%2 benchmarksVerified50.6
ReasoningWeight 17%0 benchmarksNot measuredNot measured
KnowledgeWeight 12%0 benchmarksNot measuredNot measured
MathWeight 5%0 benchmarksNot measuredNot measured
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalRank #19 of 29Percentile 36thWeight 12%4 benchmarksMixed sources63.3
Inst. FollowingWeight 5%0 benchmarksNot measuredNot measured

Chatbot Arena performance

Scroll horizontally to inspect confidence intervals and vote counts.

Chatbot Arena Elo, confidence interval, and vote count by evaluation view
ViewEloConfidence intervalVotes
Text Overall1433±4.542,281
Coding1491±6.912,219
Math1440±12.82,211
Instruction Following1431±6.514,034
Creative Writing1392±8.36,912
Multi-turn1449±8.07,452
Hard Prompts1462±5.228,157
Hard Prompts (English)1471±6.613,584
Longer Query1452±6.218,486

Benchmark Details

Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.

Agentic5 benchmarks
Terminal-Bench 2.0Provider exact
65.8%Weighted 38%
Source: Xiaomi MiMo-V2.5Provenance: Provider exact
Claw-EvalBenchmark exact
62.3%Display only
Source: Claw-Eval leaderboardProvenance: Claw-Eval reports this model as mimo_v25 in the official 2026-05-09 leaderboard snapshot. BenchLM stores the primary Pass^3 value on the local Claw-Eval display key.
MM-ClawBenchProvider exact
23.8%Display only
Source: Xiaomi MiMo-V2.5Provenance: Xiaomi reports Claw-Eval Multimodal at 23.8 on the MiMo-V2.5 launch page. BenchLM maps that to MM-ClawBench.
Gert LabsBenchmark exact

Gert Labs Composite Game Benchmark

46.89%Display only
Source: Gert Labs rankingsProvenance: Gert Labs reports this composite leaderboard score in the public rankings API. BenchLM scales the source gscore from 0-1 to 0-100 and stores it as a display-only agentic benchmark.
ResearchClawBenchBenchmark exact
16.9%Display only
Source: ResearchClawBench leaderboardProvenance: ResearchClawBench reports this model as ResearchHarness (MiMo-V2.5) in the official Pass@1 leaderboard. BenchLM stores the one-decimal RADS average on the local ResearchClawBench display key and excludes it from weighted rankings.
Coding2 benchmarks
SWE-bench ProProvider exact
56.1%Weighted 10%
Source: Xiaomi MiMo-V2.5Provenance: Provider exact
Terminal-Bench 2.0Provider exact
65.8%Display only
Source: Xiaomi MiMo-V2.5Provenance: Provider exact
Multimodal4 benchmarks
MMMU-ProProvider exact

Massive Multi-discipline Multimodal Understanding Pro

77.9%Weighted 45%
Source: Xiaomi MiMo-V2.5Provenance: Provider exact
CharXivProvider exact

CharXiv Reasoning

81%Weighted 25%
Source: Xiaomi MiMo-V2.5Provenance: Provider exact
Video-MME (with subtitle)Provider exact

Video-MME with subtitle

87.7%Display only
Source: Xiaomi MiMo-V2.5Provenance: Provider exact
Design Arena WebsiteReported

Design Arena Website Elo

1291Display only
Source: OpenRouter model benchmarksProvenance: Display-only Design Arena Website Elo synced from OpenRouter model benchmark metadata. It is excluded from BenchLM weighted scoring.

MiMo-V2.5 Family

Base entry

Related Earlier Model

MiMo-V2-Omni

Frequently Asked Questions

How does MiMo-V2.5 perform overall in AI benchmarks?

MiMo-V2.5 has 11 published benchmark scores on BenchLM, but it does not yet have enough non-generated coverage to receive a global overall rank.

Is MiMo-V2.5 good for coding and programming?

MiMo-V2.5 ranks #60 out of 122 models in coding and programming benchmarks with an average score of 50.6. There are stronger options in this category.

Is MiMo-V2.5 good for agentic tool use and computer tasks?

MiMo-V2.5 ranks #28 out of 119 models in agentic tool use and computer tasks benchmarks with an average score of 53.8. There are stronger options in this category.

Is MiMo-V2.5 good for multimodal and grounded tasks?

MiMo-V2.5 ranks #19 out of 29 models in multimodal and grounded tasks benchmarks with an average score of 63.3. There are stronger options in this category.

Which sibling models are related to MiMo-V2.5?

MiMo-V2.5 belongs to the MiMo-V2.5 family. Related variants on BenchLM include MiMo-V2.5-Pro.

Does MiMo-V2.5 have full benchmark coverage on BenchLM?

Not yet. MiMo-V2.5 currently has 11 published benchmark scores out of the 323 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.

What is the context window size of MiMo-V2.5?

MiMo-V2.5 has a published context window of 1M, which determines how much text it can process in a single interaction.

Last updated: July 21, 2026 · Runtime metrics stay blank until BenchLM has a sourced snapshot.

Choose with this week’s evidence

Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.

Free. One email per week.