XiaomiMiMo-V2.5-Pro
MiMo-V2.5-Pro is a Xiaomi open-weights model, released 23 Apr 2026. Its best configuration (thinking reasoning effort) ranks #191 of 785 on the BenchLeader Index at 57.0 ±9.4, in the upper half. The ± is the point: 418 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in instruction following (77) and lowest in reasoning (47). At $0.563 per million tokens blended it is mid-priced. Output speed of 52 tokens per second puts it in the slowest quarter, with a first answer in 39.7 s. It has been measured at 3 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 22 Apr 2026.stale: no new result in six months
- Blended price
- $0.563/M
- $0.410 in · $1.02 out
- Output speed
- 52 tok/s
- measured by Artificial Analysis
- First answer
- 40 s
- first token 1.58 s
- Context
- 1.0M
- Full answer
- 49 s
- median, reasoning included
- Cost per run
- $0.140
- one full Intelligence Index run
- Released
- 23 Apr 2026
- Overall index57
- Reasoning47
- Agents & tools49
- Knowledge63
- Instruction following77
- Long context65
- Composite59
Versions
Xiaomi has shipped 2 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| MiMo-V2.6-Pro | 21 Sept 2026 | 64.0 | #57 |
| MiMo-V2.5-Prothinkingthis page | 23 Apr 2026 | 57.0 | #191 |
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | 49.1 | #409 | 52 tok/s | 1.21 s | $0.0005 | Agents & tools 65 · Composite 50 · Instruction following 45 · Knowledge 44 · Long context 45 · Reasoning 43 |
| thinkingbest | 57.0 | #191 | 52 tok/s | 40 s | $0.0005 | Agents & tools 49 · Composite 59 · Instruction following 77 · Knowledge 63 · Long context 65 · Reasoning 47 |
| not stated | 54.9 | #239 | 36 tok/s | 1.58 s | $0.0013 | Agents & tools 40 · Coding 57 · Human preference 61 · Knowledge 54 · Maths 58 · Reasoning 66 |
How its index has moved
8 Sept 2026 to 21 Oct 2026- thinking
- not stated
- no reasoning
The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Arena Hard Promptseffort not stated | – | – | 1494#44 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 76.2%#239 | 86.6%#109 | – | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 14.8%#256 | 35.7%#116 | – | Artificial Analysis | |
| GPQA Diamond (Vals)not in indexeffort not stated | – | – | 82.6%#59 | Vals AI | |
| CritPt | 1.1%#223 | 4.0%#168 | – | Artificial Analysis | |
| LMCAnot in indexeffort not stated | – | – | 29.5%#137 | Epoch AI Benchmarking Hub | |
| DTBenchnot in indexeffort not stated | – | – | 84.5%#107 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| SciCodeeffort not stated | – | 50.2%#106 | SciCode | |
| Arena Codingeffort not stated | – | 1521#35 | Arena (LMArena) | |
| Arena WebDeveffort not stated | – | 1481#65 | Arena (LMArena) | |
| SciCode (AA)not in index | 50.6%#103 | – | Artificial Analysis | |
| LiveCodeBencheffort not stated | – | 81.3%#61 | Vals AI | |
| SWE-bench (Vals)not in indexeffort not stated | – | 74.0%#48 | Vals AI | |
| Code Migrationnot in indexeffort not stated | – | 21.6%#49 | Vals AI | |
| Vibe Code Bench v1.1not in indexeffort not stated | – | 34.1%#71 | Vals AI | |
| ALE-Benchnot in indexeffort not stated | – | 899.8#66 | Epoch AI Benchmarking Hub | |
| Terminal-Bench 4.0 (AA)not in index | 0.0%#152 | – | Artificial Analysis | |
| Terminal-Bench 2.1 (AA)not in index | 65.2%#82 | – | Artificial Analysis |
Agents & tools
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Arena Agenteffort not stated | – | – | -8.2#45 | Arena (LMArena) | |
| Terminal-Bench Hard | 35.6%#74 | 43.2%#41 | – | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 72.5%#149 | 94.2%#33 | – | Artificial Analysis | |
| Terminal-Bench 2.1 (Vals)effort not stated | – | – | 57.3%#43 | Vals AI | |
| τ³-Banking (AA) | – | 9.9%#144 | – | Artificial Analysis | |
| ITBench SRE (AA)not in index | – | 38.2%#23 | – | Artificial Analysis | |
| Analyst Agent (AA)not in index | – | 20.0%#34 | – | Artificial Analysis | |
| APEX-Agents (AA)not in index | – | 2.4%#31 | – | Artificial Analysis | |
| Finance Agent v2not in indexeffort not stated | – | – | 41.5%#59 | Vals AI | |
| Harvey's Legal Agent Benchmark v1not in indexeffort not stated | – | – | 2.1%#47 | Vals AI | |
| Legal Research Benchnot in indexeffort not stated | – | – | 15.9%#61 | Vals AI | |
| Tax Agent Benchnot in indexeffort not stated | – | – | 11.5%#56 | Vals AI | |
| Terminal-Bench 4.0 (Vals)not in indexeffort not stated | – | – | 0.5%#49 | Vals AI | |
| AutomationBenchnot in index | – | 13.7%#147 | – | Artificial Analysis | |
| GDP.pdf (AA)not in index | – | 4.0%#170 | – | Artificial Analysis |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| ProofBench | 22.0%#48 | Vals AI | |
| Arena Maths | 1479#38 | Arena (LMArena) |
Knowledge
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| AA-Omniscience | -37.8#316 | 3.3#116 | – | Artificial Analysis | |
| MMLU-Proeffort not stated | – | – | 84.6%#60 | Vals AI | |
| LegalBencheffort not stated | – | – | 77.1%#106 | Vals AI | |
| CorpFineffort not stated | – | – | 61.4%#53 | Vals AI | |
| TaxEvaleffort not stated | – | – | 73.8%#45 | Vals AI | |
| Excel Modeling Benchmarknot in indexeffort not stated | – | – | 55.2%#49 | Vals AI | |
| MedCodenot in indexeffort not stated | – | – | 32.5%#86 | Vals AI | |
| MedScribenot in indexeffort not stated | – | – | 83.7%#41 | Vals AI | |
| AA-Omniscience: accuracynot in index | 27.2%#207 | 22.4%#276 | – | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 10.8%#404 | 75.3%#22 | – | Artificial Analysis |
Instruction following
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| IFBench | 42.7%#235 | 79.9%#9 | – | Artificial Analysis | |
| Arena Instruction Followingnot in indexeffort not stated | – | – | 1467#39 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1467#51 | Arena (LMArena) | |
| EQ-Bench 4 | 1208#14 | EQ-Bench | |
| Arena Creative Writingnot in index | 1433#67 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1480#41 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1486#34 | Arena (LMArena) |
Long context
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| AA-LCR | 41.7%#352 | 79.7%#88 | Artificial Analysis | |
| MLCRnot in index | – | 5.6%#139 | Artificial Analysis |
Composite
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| AA Intelligence Index v4.3.2 | 18.3#255 | 26.0#159 | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| AtlasCloud | 103 tok/s | 4.56 s | $0.435 | $0.870 | 1.0M | fp8 |
| DigitalOcean | 29 tok/s | 0.95 s | $0.480 | $1.80 | 262k | – |
Price history
The price above, per 1M tokens, every day it changed.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.200 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0005 | $0.0004 | 45.5 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0055 | $0.0036 | 51.3 s |
| Code edit | 6,000 / 1,500 | $0.0040 | $0.0030 | 1.1 min |
| Agentic coding session | 60,000 / 4,000 | $0.029 | $0.019 | 1.9 min |
| Structured extraction | 2,000 / 200 | $0.0010 | $0.0007 | 43.6 s |
See also
Questions people ask
- What are MiMo-V2.5-Pro's benchmark scores?
- As of 24 Oct 2026, MiMo-V2.5-Pro (thinking effort) scores 57.0 ±9.4 on the BenchLeader Index, strongest in instruction following (77) and weakest in reasoning (47). That rests on 7 independent benchmark results, and every score on this page links to its publisher.
- How does MiMo-V2.5-Pro rank?
- MiMo-V2.5-Pro (thinking effort) is #191 of 785 ranked configurations on 24 Oct 2026. 418 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does MiMo-V2.5-Pro cost per 1M tokens?
- As of 24 Oct 2026, MiMo-V2.5-Pro costs $0.410 per 1M input tokens and $1.02 per 1M output tokens, $0.563 blended (three input tokens per output token). The list price last changed on 23 Oct 2026.
- When was MiMo-V2.5-Pro released?
- Xiaomi released MiMo-V2.5-Pro on 23 Apr 2026. BenchLeader has results for it at 3 reasoning-effort settings.
Data as of 24 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “MiMo-V2.5-Pro: benchmarks, pricing, speed and rank”, data as of 24 Oct 2026, https://www.benchleader.com/models/mimo-v2-5-pro.