[whatmodel.sh](../../README.md) / MiMo-V2.5 (310B-A15B) # MiMo-V2.5 (310B-A15B) Xiaomi · released 2026-04 · 310B params (15B active) · mit **score 68.0** ±0.9 · #99 of 732 · from 23 benchmark results across 6 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks MiMo-V2.5 (310B-A15B) among every model on that task or test, worst to best. 
task                vs the others          rank  counts  tests
agent tasks         ▁▁▃▆█▄▅▅●▄▂▁     #69 of 313     25%      8
coding              ▁▁▁▃▅▇█▇▅●▃▁     #74 of 469     23%      4
chat & writing      ▁▁▁▃▆▄▆█▆▇●▃    #77 of 402*    7.7%      1
math                ▁▁▃▅█▆●█▄▂▁▁   #176 of 431*    6.7%      2
vision & documents  ▁▁▂▂▁▃▄▂▆●█▄     #66 of 153     15%      4
agent tasks · #69 of 313 · 25% of its score Beats [LongCat 2.0 (1.6T-A48B)](longcat-2-0.md), [GPT-5 Pro](gpt-5-pro.md), [GPT-5](gpt-5.md) here, though they rank higher overall. 
test                                result  vs the others in its run  difficulty            counts  why
SWE-bench Pro (split unspecified)…   56.1%  ▂·▁▁▁▁▂▃●▇▂▁ beats 35/58  very hard · avg 9.9%    4.7%  sharp · at its level · small run
Agents' Last Exam overall full-pa…    8.6%  ▄···●▄▄▄█▄·█ beats 2/10   very hard · avg 1.9%    4.3%  typical
Terminal-Bench Hard                  41.7%  ▄········██● beats 5/5    very hard · avg 4.6%    4.1%  sharp · at its level
APEX-Agents (pass@1)                 13.4%  ▄▄····▄···●▄ beats 4/5    very hard · avg 1.9%    3.7%  typical
Terminal-Bench 2.0 (maker)           65.8%  ▂▂▅▅████●▂▃▃ beats 31/36  hard · avg 16%          3.1%  at its level · small run
MCP-Atlas (public set)               63.9%  ▄▄▄·······█● beats 5/5    hard · avg 19%          2.5%  at its level
SWE-bench Verified                   73.0%  ▄··········● beats 4/5    hard · avg 15%          2.0%  sharp · at its level · noisy
τ³-bench Banking                      8.7%  █·▄●······▄▄ beats 3/5    very hard · avg 5.3%    0.9%  blunt
coding · #74 of 469 · 23% of its score Beats [DeepSeek V4 Flash Vision (Exp)](deepseek-v4-flash-vision-exp.md), [GPT-5.2](gpt-5-2.md), [GPT-5](gpt-5.md) here, though they rank higher overall. Trails [Llama 3.1 Nemotron Ultra 253B v1](llama-3-1-nemotron-ultra-253b.md) here, though they rank lower overall. 
test                                result  vs the others in its run   difficulty            counts  why
Arena WebDev (Code Arena)           1438.3  ▁▁▃▂▅●▃▆▅▅▁▂ beats 59/129  avg 981                  14%  sharp
SWE-bench Pro (split unspecified)…   56.1%  ▂·▁▁▁▁▂▃●▇▂▁ beats 35/58   very hard · avg 9.9%    4.7%  sharp · at its level · small run
LiveCodeBench v6 (window unspecif…   89.1%  ▄······▄▄●·█ beats 3/5     medium · avg 50%        2.6%  typical
SWE-bench Verified                   73.0%  ▄··········● beats 4/5     hard · avg 15%          2.0%  sharp · at its level · noisy
chat & writing · #77 of 402 · 7.7% of its score Beats [Claude Haiku 5.5](claude-haiku-5-5.md), [LongCat 2.0 (1.6T-A48B)](longcat-2-0.md), [Muse Glimmer 30B](muse-glimmer-30b.md) here, though they rank higher overall. Trails [Gemma 4 26B-A4B](gemma-4-26b-a4b.md), [Nemotron Cascade 2 30B-A3B](nemotron-cascade-2-30b-a3b.md), [Gemma 4 31B](gemma-4-31b.md) here, though they rank lower overall. 
test            result  vs the others in its run    difficulty  counts  why
LMArena (text)  1433.9  ▁▁▁▂▃▃▃▆▆▇●▃ beats 291/385  avg 1260      7.7%  blunt
math · #176 of 431 · 6.7% of its score Trails [OpenThinker3 7B](openthinker3-7b.md), [Olmo 3 7B Think](olmo-3-7b-think.md), [OpenThinker2 32B](openthinker2-32b.md) here, though they rank lower overall. 
test                result  vs the others in its run  difficulty        counts  why
AIME 2026            92.3%  ▄▄··●····█·▄ beats 2/5    medium · avg 65%    4.4%  sharp · maxed out
HMMT February 2026   61.4%  ●▄▄········█ beats 0/5    hard · avg 31%      2.3%  sharp · noisy
vision & documents · #66 of 153 · 15% of its score Beats [DeepSeek V4 Flash Vision (Exp)](deepseek-v4-flash-vision-exp.md), [GPT-5](gpt-5.md), [Muse Glimmer 30B](muse-glimmer-30b.md) here, though they rank higher overall. Trails [Gemma 4 26B-A4B](gemma-4-26b-a4b.md), [Qwen3.6 Flash](qwen3-6-flash.md), [Qwen3.5 35B-A3B](qwen3-5-35b-a3b.md) here, though they rank lower overall. 
test                                result  vs the others in its run   difficulty        counts  why
Arena vision (style controlled)     1233.8  ▁▁▂▂▂▂▃▄▅●█▇ beats 92/152  avg 1061            5.1%  blunt
CharXiv (reasoning questions) (ma…   81.0%  ▂▅▃▅▅▆▆▅▆●▆▂ beats 82/101  medium · avg 44%    5.1%  small run · several runs
SciDocBench scientific-document w…   40.2%  ●▃···█▃▅▃·▃▅ beats 0/11    hard · avg 25%      2.6%  blunt · at its level
MMMU Pro (maker)                     77.9%  ▂▃▁▃▆▇▃▇▇█●▅ beats 46/52   medium · avg 48%    2.0%  at its level · small run
other tests (8) 
test                               result  vs the others in its run  difficulty            counts  why
GPQA Diamond                        83.0%  █···██·●··██ beats 3/5    hard · avg 41%           13%  sharp
Humanity's Last Exam                24.3%  █▄······●·▄▄ beats 3/5    very hard · avg 4.7%    5.9%  typical
MMLU-Pro                            84.6%  ▄·▄▄····●··█ beats 3/5    medium · avg 55%        4.3%  typical
IFBench (aggregation unspecified)   67.1%  ▄···●▄··▄··█ beats 1/5    medium · avg 43%        2.9%  at its level
MultiChallenge                      39.0%  ●··▄·▄·····█ beats 0/5    hard · avg 25%          2.5%  at its level
Daily-Omni (maker)                  83.5%  █··········● beats 1/1                               —  not counted yet: no other model in the …
HR-Bench 4K (maker)                 88.5%  █··········● beats 1/1                               —  not counted yet: no other model in the …
Video-Holmes (maker)                64.0%  █··········● beats 1/1                               —  not counted yet: no other model in the …
## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md) 
source                                    results  share of the score
Solar Open 2 Technical Report                  13   51% ▇▇▇▇▇▇
LMArena leaderboard dataset                     3   26% ▇▇▇
XiaomiMiMo/MiMo-V2.5 model card (submit…        4   12% ▇
Agents' Last Exam                               1    4% ▇
SciDocBench: A Workflow-Centered Benchm…        1    3% ▇
Xiaomi MiMo vendor-reported benchmark s…        1    3% ▇
prior: pull toward the middle                        0% ▇
weighted average                                   = 68.0
Licences: Solar Open 2 Technical Report ([CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · XiaomiMiMo/MiMo-V2.5 model card ([MIT](../../licenses/XiaomiMiMo--MiMo-V2.5.txt)) · Agents' Last Exam ([CC BY 4.0](../../licenses/paper-2606.05405v2-CC-BY-4.0.txt)) · SciDocBench: A Workflow-Centered Benchmark and Data Pipeline for Scientific Document Understanding ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/))
Not counted (3) 
source                  test          score  why
XiaomiMiMo/MiMo-V2.5 …  Daily-Omni    83.5%  too few models have it yet
XiaomiMiMo/MiMo-V2.5 …  HR-Bench 4K   88.5%  too few models have it yet
XiaomiMiMo/MiMo-V2.5 …  Video-Holmes  64.0%  too few models have it yet
## How to run it ### Through an API api (Novita AI) · $0.21/M blended · $0.168 in / $0.336 out per 1M tokens · [novita.ai](https://novita.ai/models/llm/xiaomimimo-mimo-v2.5) [price source](https://novita.ai/pricing): USD per 1M tokens, Novita AI serverless chat completions, model xiaomimimo/mimo-v2.5 (status active; 1M context): input $0.168, output $0.336; cache read $0.0034 excluded. Xiaomi's own API lists mimo-v2.5 at $0.14/$0.28 (overseas, cache miss) but marks it 'to be deprecated' on 2026-10-21. Currency: USD. Novita AI official pricing page, observed 2026-09-25. · [context source](https://github.com/anomalyco/models.dev/blob/a668908cfe7073faa0363699486fc41a3ff52412/models/xiaomi/mimo-v2.5.toml) ### llama.cpp (recommended) 1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md) ```sh # macOS / Linux curl -LsSf https://llama.app/install.sh | sh # Windows irm https://llama.app/install.ps1 | iex ``` 2. Run MiMo-V2.5 (310B-A15B) (ud-q4_k_xl, needs ~186.5 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/MiMo-V2.5-GGUF/tree/main)) ```sh llama cli -hf unsloth/MiMo-V2.5-GGUF:UD-Q4_K_XL ``` 3. Or start a local server and open the address it prints (http://localhost:9931) ```sh llama serve -hf unsloth/MiMo-V2.5-GGUF:UD-Q4_K_XL ``` Other versions: ud-q2_k_xl (~93.5 GB) `llama cli -hf unsloth/MiMo-V2.5-GGUF:UD-Q2_K_XL` ### LM Studio 1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download) 2. Run MiMo-V2.5 (310B-A15B) (ud-q4_k_xl, needs ~186.5 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/MiMo-V2.5-GGUF/tree/main)) ```sh lms get https://huggingface.co/unsloth/MiMo-V2.5-GGUF@UD-Q4_K_XL -y ``` 3. Then run lms ls to see its name, and lms chat followed by that name Other versions: ud-q2_k_xl (~93.5 GB) `lms get https://huggingface.co/unsloth/MiMo-V2.5-GGUF@UD-Q2_K_XL -y` Model repository and documentation: [huggingface.co/XiaomiMiMo](https://huggingface.co/XiaomiMiMo/MiMo-V2.5) [interactive version](https://whatmodel.sh/model/mimo-v2-5) · [JSON](https://whatmodel.sh/api?format=json&all=1)