MiMo V2.5
MiMo V2.5 is capable in multimodal tasks and long context; and behind the leaders in reasoning, instruction following, coding, and agentic tasks. Too few results yet to rate safety, math, multilingual tasks, or factuality.
Capability profile
Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.
Too few results yet to rate Safety, Math, Multilingual or Factuality.
Price
$0.11input$0.22outputper million tokens
From Artificial Analysis · 2 providers tracked · All prices
Evidence
68results on54benchmarks
- 3 independently verified
- 17 aggregator
- 25 vendor-reported
- 23 cross-referenced
From 8 sources · latest Nov 13, 2026 · How verification works
API features
Tool callingStructured outputsReasoning
As listed by OpenRouter
Research
3 papers reference MiMo V2.5MiMo V2.5 benchmark results
68 results on 54 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.
22.9% behind the leader3 of 6 ranked benchmarks measured
- 87.70Nov 13, 2026
- CharXiv (reasoning)81.00Nov 13, 2026CharXiv-R
- MMMU-Pro75.43Oct 16, 2026aa_mmmu_pro
Show 4 more multimodal resultsHide 4 multimodal results
- 1247.51Aug 25, 2026
- 1237Jun 17, 2026
- 77.90Nov 13, 2026
- MMMU-Pro75.40Aug 24, 2026MMMU Pro (Standard 10)
24.4% behind the leader1 of 3 ranked benchmarks measured
- 73.00Oct 16, 2026
29.9% behind the leader3 of 6 ranked benchmarks measured
- GPQA Diamond84.95Oct 16, 2026gpqa
- Humanity's Last Exam27.20Oct 16, 2026aa_hle
- 3.71Oct 16, 2026
Show 5 more reasoning resultsHide 5 reasoning results
- 4.00Jul 30, 2026
- 3.70Jul 30, 2026
- 58.10Oct 29, 2026
- 84.90Jul 30, 2026
- Humanity's Last Exam25.20Jul 30, 2026HLE text only
30.6% behind the leader1 of 3 ranked benchmarks measured
- IFBench67.14Oct 16, 2026aa_ifbench
Show 1 more instruction following resultHide 1 instruction following result
- 67.10Aug 24, 2026
35.7% behind the leader7 of 10 ranked benchmarks measured
- 1437.71May 22, 2026
- Terminal-Bench 2.163.67Oct 16, 2026terminalbenchV21
- SciCode43.87Oct 16, 2026aa_scicode
- Terminal-Bench Hard41.67Oct 16, 2026aa_terminalbench_hard
- 56.10Nov 13, 2026
- 35.50Oct 29, 2026
- SWE-bench Verified30.80Oct 29, 2026SWE-Bench (AgentLess)
Show 3 more coding resultsHide 3 coding results
- 43.10Jul 30, 2026
- 71.00Jul 30, 2026
- 63.70Jul 30, 2026
47.7% behind the leader3 of 7 ranked benchmarks measured
- 24.86Oct 16, 2026
- τ-Bench V3 · Banking8.66Oct 16, 2026tauBanking
- Terminal-Bench 4.00.00Oct 16, 2026
0 of 5 ranked benchmarks measured
- 82.60Jul 30, 2026
- 93.60Jul 30, 2026
0 of 4 ranked benchmarks measured
- AA-Omniscience · Accuracy16.75Oct 16, 2026omniscienceAccuracy
- AA-Omniscience · Non-hallucination68.07Oct 16, 2026omniscienceNonHallucination
- 16.10Aug 24, 2026
More results
Benchmarks outside the capability baskets. They are not ranked against a leader.
- AA Intelligence21.82Oct 16, 2026aa_intelligence_index
- τ²-Bench Telecom (AA run)90.64Oct 16, 2026aa_tau2
- -9.83Oct 16, 2026
- 36.73Nov 13, 2026
- 64.00Nov 13, 2026
- 87.20Nov 13, 2026
Show 26 more resultsHide 26 results
- aime_2024_2025_combined36.90Oct 29, 2026AIME 24&25
- 96.50Oct 29, 2026
- Claw Eval (pass@3)63.20Nov 13, 2026Claw-Eval
- 88.20Oct 29, 2026
- 83.70Oct 29, 2026
- 1265.10Jul 30, 2026
- 1145.00Jul 30, 2026
- 83.50Aug 24, 2026
- 77.40Oct 29, 2026
- 88.60Oct 29, 2026
- HLE (with tools)40.00Jul 30, 2026HLE with tools
- 67.70Oct 29, 2026
- 70.90Oct 29, 2026
- 73.60Aug 24, 2026
- 86.30Oct 29, 2026
- 65.80Oct 29, 2026
- 89.80Oct 29, 2026
- 99.30Aug 24, 2026
- 56.10Jul 30, 2026
- 65.80Nov 13, 2026
- 49.10Jul 30, 2026
- 80.70Oct 29, 2026
- 86.40Aug 24, 2026
- 84.70Oct 29, 2026
- τ³-Bench Banking9.00Jul 30, 2026τ³-Banking
- τ³-Bench Banking6.60Jul 30, 2026Tau 3 Banking
MiMo V2.5: common questions
Who makes MiMo V2.5?
MiMo V2.5 is made by Xiaomi.
When was MiMo V2.5 released?
MiMo V2.5 was released on Apr 22, 2026, according to Artificial Analysis.
What is MiMo V2.5 good at?
MiMo V2.5 is capable in multimodal tasks and long context; and behind the leaders in reasoning, instruction following, coding, and agentic tasks. Too few results yet to rate safety, math, multilingual tasks, or factuality.
How much does MiMo V2.5 cost?
MiMo V2.5 costs $0.11 per million input tokens and $0.22 per million output tokens, according to Artificial Analysis. We track its price at 2 providers. At a mix of three input tokens to one output token, it is cheaper than 87% of the 331 priced models we track.
How many benchmarks has MiMo V2.5 been tested on?
We track 68 results for MiMo V2.5 on 54 benchmarks from 8 sources, 3 of them independently verified. The latest was recorded on Nov 13, 2026.
Which API features does MiMo V2.5 support?
OpenRouter lists tool calling, structured outputs, and reasoning for MiMo V2.5.
About this record
Where MiMo V2.5's numbers come from, and every name it appears under.
- Tracked since
- Sep 23, 2026
- Newest source mention
- Sep 23, 2026
Where the results come from
Verification: 68 scores · 3 independently verified · 17 aggregator-attributed · 23 vendor cross-reference · 25 vendor-reported. How these tiers are assigned
From 8 sources on 6 sites. Hugging Face supplies 26 of them; the 3 independently verified results come from 2 sites. Bars are coloured by trust tier.
- huggingface.co26
- artificialanalysis.ai17
- thinkingmachines.ai13
- api.llm-stats.com9
- datasets-server.huggingface.co2
- lmarena.ai1