Compare AI models: DeepSeek V4 Pro vs MiMo-V2.5

ReasoningCodingCost efficiencyContext windowMultimodalTool use

Normalized model capability scores
Model	Reasoning	Coding	Cost efficiency	Context window	Multimodal	Tool use
DeepSeek V4 Pro	100/100	100/100	33/100	82/100	0/100	68/100
MiMo-V2.5	17/100	53/100	76/100	93/100	67/100	50/100

Overview

Author

Context length

Reasoning

True

Input modalities

Text

Text, Image, Audio, Video

Output modalities

Text

Providers

Pricing

Input

$1.60/ 1M

$0.14/ 1M

Output

$3.20/ 1M

$0.28/ 1M

Cached input

$0.14/ 1M

$0.0028/ 1M

MomentumLast 2 mo

Unique users

450K

1.2M

Completed sessions

1,400,457

47,982,179

Token share

0.49%

2.89%

Tokens

5.6T-82%

33T+1%

AUG 30OCT 24

RetentionWeek 1

Returning users

58.3%44K user-weeks

70.6%23K user-weeks

Related comparisons. Other model pairs to check.