MiMo v2.5
MiMo v2.5Rankingin thinking mode?
Alternatives?
CheaperAgnes 2.5 Pro Beta+61%-14%$0.10 / $0.30
BetterGPT-5.6 Lunamedium+15%-18%$0.016
OpenDeepSeek V4.1 Flash 1017max×2.1+1,274%$0.26Memory (4-bit): 186 GB → 458 GB+5CheaperGPT-6 Lunamedium+37%-8%$0.017Function calling only with none · Chat Completions+3
CheaperLing 3.0 Flash VLthinkingOpen+13%-49%$0.06 / $0.18Context: 1M → 262KAgnes 2.5 Pro Beta+61%-14%$0.10 / $0.30GPT-6 Lunamedium+37%-8%$0.017Function calling only with none · Chat Completions+3GPT-5.6 Lunamedium+15%-18%$0.016Ling 3.0 Flash VLthinkingOpen+13%-49%$0.06 / $0.18Context: 1M → 262KLing 3.0 Flash FinthinkingOpensame-49%$0.06 / $0.18Context: 1M → 262K+1GPT-5.6 Lunamedium+15%-18%$0.016GPT-6 Lunamedium+37%-8%$0.017Function calling only with none · Chat Completions+3
Claude Haiku 5.5low+35%+28%$0.024Remove sampling parameters+5MiMo v2.5 ProthinkingOpen+19%+184%$0.054Memory (4-bit): 186 GB → 614 GB+1MiMo v2.6 FlashthinkingOpen+74%+227%$0.062Pass back reasoning_content+3DeepSeek V4.1 Flash 1017maxOpen×2.1+1,274%$0.26Memory (4-bit): 186 GB → 458 GB+5
GLM 5.3maxOpen×2.1+10,417%$2.01Memory (4-bit): 186 GB → 452 GB+5
Step 5 PreviewthinkingOpen×2+5,282%$1.03Memory (4-bit): 186 GB → 363 GB+3GLM 5.3 FlashmaxOpen+92%+1,387%$0.28Memory (4-bit): 186 GB → 193 GB+5Ling 3.1 FlashOpen+88%+5,142%$1.00Memory (4-bit): 186 GB → 343 GB+1MiMo v2.6 ProthinkingOpen×2.1+599%$0.13Memory (4-bit): 186 GB → 615 GB+4MiMo v2.6 FlashthinkingOpen+74%+227%$0.062Pass back reasoning_content+3MiMo v2 Prothinking+31%—$0.43 / $0.87No image inputMiMo v2.5 ProthinkingOpen+19%+184%$0.054Memory (4-bit): 186 GB → 614 GB+1MiMo v2 Omni 0327+15%——Context: 1M → 256KMiMo v2 Omnithinking+10%—$0.14 / $0.28Context: 1M → 262KMiMo v2.6 Pro UltraSpeedOpenn/a—$4.35 / $8.70Memory (4-bit): not verifiedMiMo VL 7B RL 2508Openn/a——Context: 1M → 128KAll Xiaomi models →Run it yourself
estimateBenchmarks
all categoriesFAQ
How much does MiMo v2.5 cost?›
$0.14 per 1M input tokens and $0.28 per 1M output tokens; cached input $0.003 per 1M. A typical benchmark task costs about $0.019 at default effort.
Is MiMo v2.5 free to use?›
There is no free API; the cheapest paid price is $0.14 per 1M input tokens. The weights are open, so you can run it yourself.
How good is MiMo v2.5, and what is it best at?›
#98 of 334 in Chat & Knowledge, #75 of 132 in Coding, #62 of 133 in Agents & Tool Use, #72 of 195 in Writing & Language, #44 of 124 in Vision & Documents. Strongest at: Tool calling.
Is MiMo v2.5 better than G9v3 39A5B?›
On AA Intelligence Index 4.3.2, MiMo v2.5 scores 21.8 vs 21.8 — about the same.
How can I use MiMo v2.5?›
Through the API of DeepInfra, Novita, Venice, Xiaomi.
What is the context window of MiMo v2.5?›
1M tokens, with up to 131K output tokens.
When was MiMo v2.5 released? Is there a newer version?›
Released 2026-04-22. It has been superseded by MiMo v2.6 Pro.
Is MiMo v2.5 open source?›
Yes, the weights are open (MIT licence).
Can I run MiMo v2.5 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can MiMo v2.5 read images?›
It accepts text and images, audio, video as input.
What is MiMo v2.5?›
MiMo v2.5 is an open-weights language model from Xiaomi, released Apr 2026.
What are the best alternatives to MiMo v2.5?›
Agnes 2.5 Pro Beta (cheaper), GPT-5.6 Luna (better), DeepSeek V4.1 Flash 1017 (open weights).