ModelCap
Current AI model comparison
Exact identities · public evidence
DeepSeek
DeepSeek V4.1 Flash
Position
#40
Index
65.2
Estimated
publisher-corpus-prior over 185 held-out anchors (22%); opens from measured predecessor deepseek/deepseek-v4-flash-0731 (78%); shrunk 2 toward the measured corpus
Comparison side 1
Xiaomi
MiMo-V2.5
Position
#41
Index
64.5
Measured
2 public benchmark observations across 2 boards
Comparison side 2
Output / 1M
$0.60
vs $0.28
Context
1M
vs 1M
Weights
Open weights
vs Open weights