Geekbench AI Score

18503
Single Precision Score
32495
Half Precision Score
31504
Quantized Score
Geekbench AI 1.7.0 for macOS AArch64

Result Information

User StefanGentz
Upload Date March 18 2026 10:58 PM
Views 11

System Information

AI Information
Framework Core ML
Backend GPU
Device Apple M5 Pro
System Information
Operating System macOS 26.3.1 (Build 25D2128)
Model Mac17,9
Model IDMac17,9
Motherboard Mac17,9
CPU Information
Name Apple M5 Pro
Topology 1 Processor, 15 Cores
Identifier Apple M5 Pro
Base Frequency 4.09 GHz
Cluster 15 Cores
Cluster 210 Cores
Memory Information
Size 48.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 8270
1.54 KIPS
 
Image Classification (HP)
100% 12117
2.25 KIPS
 
Image Classification (Q)
99% 11237
2.10 KIPS
 
Image Segmentation (SP)
100% 15372
249.2 IPS
 
Image Segmentation (HP)
100% 24992
405.1 IPS
 
Image Segmentation (Q)
99% 23918
387.7 IPS
 
Pose Estimation (SP)
100% 60064
70.1 IPS
 
Pose Estimation (HP)
100% 156115
182.2 IPS
 
Pose Estimation (Q)
95% 190298
223.0 IPS
 
Object Detection (SP)
100% 8043
638.0 IPS
 
Object Detection (HP)
100% 14115
1.12 KIPS
 
Object Detection (Q)
94% 13088
1.04 KIPS
 
Face Detection (SP)
100% 30497
362.4 IPS
 
Face Detection (HP)
100% 50996
605.9 IPS
 
Face Detection (Q)
100% 45214
537.2 IPS
 
Depth Estimation (SP)
100% 39362
303.3 IPS
 
Depth Estimation (HP)
99% 70130
540.3 IPS
 
Depth Estimation (Q)
86% 64215
501.3 IPS
 
Style Transfer (SP)
100% 129424
166.4 IPS
 
Style Transfer (HP)
100% 284759
366.1 IPS
 
Style Transfer (Q)
96% 334094
431.1 IPS
 
Image Super-Resolution (SP)
100% 22290
823.1 IPS
 
Image Super-Resolution (HP)
100% 40596
1.50 KIPS
 
Image Super-Resolution (Q)
99% 33233
1.23 KIPS
 
Text Classification (SP)
100% 3047
4.07 KIPS
 
Text Classification (HP)
99% 5544
7.40 KIPS
 
Text Classification (Q)
93% 5463
7.33 KIPS
 
Machine Translation (SP)
100% 7259
125.0 IPS
 
Machine Translation (HP)
100% 8585
147.9 IPS
 
Machine Translation (Q)
98% 8170
141.2 IPS
 

Compare

Set Baseline Geekbench 6 now available