MiMo-V2.5-Pro is Xiaomi’s flagship model, delivering strong performance in general agentic capabilities, complex software engineering, and long-horizon tasks, with top rankings on benchmarks such as ClawEval, GDPVal, and SWE-bench Pro. It can independently and autonomously complete professional tasks that would take human experts days or weeks, involving more than a thousand tool calls. Its context length of up to 1M makes it well suited for integration with a wide range of agent frameworks.
Reasoning|Proprietary Model
Knowledge Cutoff
Unknown
Input → Output Format
Context Memory
1.0MIN131KOUT
AI Performance Evaluation
Arena Overall Score
1463
±8As of 2026-05-01
Overall Rank
No.22
5,037 Votes
Arena by Ability
Hard Prompts
1493±11No.17
Expert Knowledge
1506±26No.8
Instruction Following
1468±15No.13
Conversation Memory
1478±21No.22
Creative
1428±22No.39
Coding
1511±17No.25
Math
1477±32No.15
Arena by Occupation
Creative Writing
1453±18No.19
Social Sciences
1501±20No.5
Media
1431±19No.34
Business
1469±19No.20
Healthcare
1492±31No.17
Legal
1488±30No.14
Software
1502±13No.21
Mathematics
1507±33No.9
Source:Arena Intelligence
Overall
AA Intelligence Index
54%↑15%
Reasoning & Math
GPQA Diamond
87%↑4%
HLE
34%↑16%
Coding
AA Coding Index
46%↑9%
TAU2
94%↑14%
TerminalBench
43%↑9%
SciCode
50%↑8%
Language & Instructions
IFBench
80%↑17%
AA-LCR
73%↑11%
Output Speed
Standard Mode
60tok/s↓18
First Output 1.98s
Source:Artificial Analysis
