Xiaomi model product
MiMo
5 is Xiaomi's native omnimodal sparse Mixture-of-Experts model with 310B total parameters, 15B activated parameters, and a 1M-token context window.
Updated Aug 12, 2026. Default version: MiMo-V2.5
LLMBoard score61.8MiMo-V2.5
Coverage20%9 benchmark families
Context window1MTokens
Official input price$0.14Xiaomi API
MiMo Specifications
Technical details for the model's default version.
- Version
- MiMo-V2.5
- Released
- Apr 22, 2026
- Knowledge cutoff
- Dec 1, 2024
- Parameters
- 310.8B
- Context window
- 1M
- Max output
- 131.1K
- Inputs
- audio, image, text, video
- Outputs
- text
- Open weights
- Yes
- License
- MIT
MiMo Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
MiMo Benchmark Results
Benchmark scores for MiMo-V2.5.
| DailyOmni | 83.5% | 01 | 1 | 100.0% | C | |
| HR-Bench (4k) | 88.5% | 01 | 1 | 100.0% | C | |
| ResearchClawBench | 16.9% | 01 | 1 | 100.0% | C | |
| GraphWalks | 87.0% | 02 | 3 | 50.0% | C | |
| MiMo Coding Bench | 71.8% | 02 | 2 | 0.0% | C | |
| OmniDocBench | 87.2% | 02 | 2 | 0.0% | C | |
| VideoHolmes | 64.0% | 03 | 3 | 0.0% | C | |
| Video-MME | 87.7% | 04 | 17 | 81.3% | C | |
| Claw-Eval | 63.2% | 07 | 13 | 50.0% | C | |
| Terminal-Bench 2.0 | 65.8% | 16 | 49 | 68.8% | C | |
| CharXiv-R | 81.0% | 19 | 48 | 61.7% | C | |
| Finance Agent v2 | 36.7% | 21 | 26 | 20.0% | B | |
| MMMU-Pro | 77.9% | 22 | 66 | 67.7% | C | |
| SWE-Bench Pro | 56.1% | 29 | 45 | 36.4% | C |
MiMo Arena Results
Preference and agent-evaluation results for the default version.
| vision | overall | 46 | 1251.4 | 20,781 | N/A | |
| webdev | overall | 47 | 1437.7 | 11,882 | N/A | |
| vision style control | overall | 50 | 1237.7 | 20,781 | N/A | |
| text factuality | overall | 77 | 1439.0 | 44,582 | N/A | |
| text | overall | 81 | 1427.4 | 44,653 | N/A | |
| text style control | overall | 86 | 1433.8 | 44,653 | N/A |
MiMo Runtime Performance
Provider-specific output speed and catalog latency for MiMo-V2.5. Runtime does not affect the capability score.
No runtime data
Browse runtime rankingsNo provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
MiMo Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.14 input, $0.28 output per 1M
- Official provider
- Xiaomi
- Lowest third-party
- From $0.14 input, $0.28 output per 1M via Impossibl
- Tracked offerings
- 21
| Kenari | mimo-v2-5 | global | N/A | N/A | 1M | |
| Xiaomi Token Plan (China) | mimo-v2.5 | global | N/A | N/A | 1M | |
| Xiaomi Token Plan (Europe) | mimo-v2.5 | global | N/A | N/A | 1M | |
| Xiaomi Token Plan (Singapore) | mimo-v2.5 | global | N/A | N/A | 1M | |
| Impossibl | xiaomi/mimo-v2.5 | global | $0.14 | $0.28 | 1M | |
| OpenCode Go | mimo-v2.5 | global | $0.14 | $0.28 | 1M | |
| NanoGPT | xiaomi/mimo-v2.5 | global | $0.14 | $0.28 | 1M | |
| LLM Gateway | mimo-v2.5 | global | $0.14 | $0.28 | 1M | |
| ClinePass | cline-pass/mimo-v2.5 | global | $0.14 | $0.28 | 1M | |
| Pioneer | XiaomiMiMo/MiMo-V2.5 | global | $0.14 | $0.28 | 1.1M | |
| Venice AI | xiaomi-mimo-v2-5 | global | $0.14 | $0.28 | 1M | |
| Kilo Gateway | xiaomi/mimo-v2.5 | global | $0.14 | $0.28 | 1M | |
| Vercel AI Gateway | xiaomi/mimo-v2.5 | global | $0.14 | $0.28 | 1.1M | |
| OpenRouter | xiaomi/mimo-v2.5 | global | $0.14 | $0.28 | 1.1M | |
| Xiaomi | mimo-v2.5 | global | $0.14 | $0.28 | 1M | |
| CrossModel | xiaomi/mimo-v2.5 | global | $0.16 | $0.32 | 1M | |
| Ambient | xiaomi/mimo-v2.5 | global | $0.40 | $2 | 1M | |
| ZenMux | xiaomi/mimo-v2.5 | global | $0.40 | $2 | 1M | |
| Hugging Face | XiaomiMiMo/MiMo-V2.5 | global | $0.40 | $2 | 262.1K | |
| Deep Infra | XiaomiMiMo/MiMo-V2.5 | global | $0.40 | $2 | 262.1K | |
| EmpirioLabs AI | mimo-v2-5 | global | $0.70 | $1.4 | 1M |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
MiMo Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| MiMo-V2.5 | 61.8 | 310.8B | 1M | 131.1K | Yes | MIT |
What is MiMo?
Key information about MiMo and its available data.
5 is Xiaomi's native omnimodal sparse Mixture-of-Experts model with 310B total parameters, 15B activated parameters, and a 1M-token context window.
Built on the MiMo-V2-Flash backbone, it adds dedicated vision and audio encoders for text, image, video, and audio understanding, and is post-trained with SFT, agentic reinforcement learning, and Multi-Teacher On-Policy Distillation for multimodal perception, long-context reasoning, and agentic workflows.
Data as of 2026-08-11.
MiMo vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to MiMo
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
XI
MiMo V2 Pro
Xiaomi
XI
MiMo V2 Omni
Xiaomi
XI
MiMo V2 Flash
Xiaomi
XI
MiMo V2.5 Pro
Xiaomi
OP
GPT-5.1
OpenAI
OP
GPT-5.1-Instant
OpenAI
FAQ
Common questions about MiMo.
When was MiMo released?
MiMo's default version was released on Apr 22, 2026.
How much does MiMo cost?
MiMo's official API price is $0.14 per million input tokens and $0.28 per million output tokens via Xiaomi. The lowest tracked third-party offer starts at $0.14 input and $0.28 output via Impossibl.
Who created MiMo?
MiMo was created by Xiaomi.
What is the context window for MiMo?
The default version has a 1M token context window.
Is MiMo open weight?
Yes. The default version is marked as open weight under MIT.
How many API providers offer MiMo?
21 provider offerings are linked to the default version.
What models should I compare MiMo with?
Nearby ranked alternatives include GPT-5.3-Codex, Nemotron 3 Ultra, GPT-5.1.