MiMo-V2-Flash

MiMo-V2-Flash is a powerful, efficient, and ultra-fast foundation language model that excels in reasoning, coding, and agentic scenarios. It is a Mixture-of-Experts model with 309B total parameters and 15B active parameters, featuring a hybrid attention architecture with sliding-window and full attention (5:1 ratio, 128-token window).

AIME 2025

94.1%

i
Arena-Hard v2

86.2%

i
MMLU-Pro

84.9%

i
HMMT 2025

84.4%

i
GPQA

83.7%

i
LiveCodeBench v6

80.6%

i
Tau-bench

80.3%

i
SWE-Bench Verified

73.4%

i
SWE-bench Multilingual

71.7%

i
LongBench v2

60.6%

i
BrowseComp

58.3%

i
MRCR

45.7%

i
Terminal-Bench 2.0

38.5%

i
Terminal-Bench

30.5%

i
Humanity's Last Exam

22.1%

i

Pricing, uptime, and speed via OpenRouter — updated Jun 17, 2026, 04:09 PM.

Provider	Status	Input	Output	Limits	Uptime	Speed	Notes
Xiaomi	available	$0.10/Mtok cache $0.01/Mtok	$0.30/Mtok	262K tokens context 66K tokens max output	100.0% 5m 100.0%	1,094 ms p50 TTFT 37 tok/s p50	fp8