ModelCap

Current AI model comparison

Exact identities · public evidence

DeepSeek

DeepSeek V4 Pro 0813

Position

#21

Index

73.5

Measured

4 public benchmark observations across 4 boards

Comparison side 1

Google

Gemini 3.8 Flash

Position

#7

Index

82.2

Measured

4 public benchmark observations across 4 boards

Comparison side 2

Output / 1M

$1.74

vs $3.75

Context

1M

vs 1M

Weights

Open weights

vs API only