GPT-5-mini 2025-08-07
OpenAI
Parameters
—
Architecture
—
Released
—
License
—
About
No detailed description available for this model.
Benchmark Scores
| Benchmark | Score | Date | Standing |
|---|---|---|---|
|
MMMU-Pro
vision_language
|
70.88%
|
01.02.2026 |
#28/35
better than 21%
|
|
MMLU-Pro
knowledge
|
76.45%
|
01.02.2026 |
#23/48
better than 53%
|
|
MMLU-Redux
knowledge
|
92.02%
|
01.02.2026 |
#6/30
better than 79%
|
|
SuperGPQA
knowledge
|
72.97%
|
01.02.2026 |
#12/28
better than 59%
|
|
IFEval
instruction_following
|
95.40%
|
01.02.2026 |
#5/26
better than 84%
|
|
Multi-Challenge
instruction_following
|
76.39%
|
01.02.2026 |
#12/18
better than 29%
|
|
LongBench v2
long_context
|
72.17%
|
01.02.2026 |
#8/22
better than 67%
|
|
Humanity's Last Exam
stem_reasoning
|
32.77%
|
01.02.2026 |
#52/71
better than 27%
|
|
GPQA Diamond
stem_reasoning
|
78.26%
|
01.02.2026 |
#57/80
better than 28%
|
|
HMMT Feb 25
stem_reasoning
|
87.56%
|
01.02.2026 |
#8/17
better than 56%
|
|
HMMT Nov 25
stem_reasoning
|
37.56%
|
01.02.2026 |
#19/20
better than 5%
|
|
SWE-bench Verified
coding_agent
|
82.11%
|
01.02.2026 |
#34/60
better than 42%
|
|
Terminal-Bench 2.0
coding_agent
|
31.90
|
01.02.2026 |
#23/23
better than 0%
|
|
LiveCodeBench v6
stem_reasoning
|
82.26%
|
01.02.2026 |
#24/49
better than 52%
|
|
CodeForces
stem_reasoning
|
60.99%
|
01.02.2026 |
#8/18
better than 59%
|
|
OJBench
stem_reasoning
|
55.51%
|
01.02.2026 |
#5/22
better than 81%
|
|
FullStackBench en
coding_agent
|
30.60
|
01.02.2026 |
#5/5
better than 0%
|
|
FullStackBench zh
coding_agent
|
35.20
|
01.02.2026 |
#5/5
better than 0%
|
|
BFCL-V4
general_agent
|
59.95%
|
01.02.2026 |
#16/27
better than 42%
|
|
TAU2-Bench
general_agent
|
66.30%
|
01.02.2026 |
#10/21
better than 55%
|
|
VITA-Bench
general_agent
|
24.61%
|
01.02.2026 |
#11/16
better than 33%
|
|
BrowseComp-zh
general_agent
|
69.37%
|
01.02.2026 |
#4/13
better than 75%
|
|
WideSearch
general_agent
|
23.48%
|
01.02.2026 |
#22/24
better than 9%
|
|
Seal-0
general_agent
|
34.20
|
01.02.2026 |
#4/4
better than 0%
|
|
MMMLU
multilingual
|
82.10%
|
01.02.2026 |
#8/20
better than 63%
|
|
MMLU-ProX
multilingual
|
49.03%
|
01.02.2026 |
#12/19
better than 39%
|
|
NOVA-63
multilingual
|
51.90
|
01.02.2026 |
#6/6
better than 0%
|
|
INCLUDE
multilingual
|
83.72%
|
01.02.2026 |
#3/11
better than 80%
|
|
Global PIQA
multilingual
|
100.00%
|
01.02.2026 |
#1/6
better than 100%
|
|
PolyMATH
multilingual
|
85.50%
|
01.02.2026 |
#3/11
better than 80%
|
|
WMT24++ (en→xx)
multilingual
|
87.82%
|
01.02.2026 |
#8/17
better than 56%
|
|
MAXIFE
multilingual
|
73.00%
|
01.02.2026 |
#4/6
better than 40%
|
|
MMMU
vision_language
|
86.90%
|
01.02.2026 |
#12/19
better than 39%
|
|
MathVision
vision_language
|
50.00%
|
01.02.2026 |
#19/27
better than 31%
|
|
DynaMath
vision_language
|
29.21%
|
01.02.2026 |
#10/14
better than 31%
|
|
ZEROBench
vision_language
|
3.00
|
01.02.2026 |
#13/14
better than 0%
|
|
ZEROBench_sub
vision_language
|
12.75%
|
01.02.2026 |
#7/11
better than 40%
|
|
VlmsAreBlind
vision_language
|
75.80
|
01.02.2026 |
#11/11
better than 0%
|
|
RealWorldQA
vision_language
|
77.91%
|
01.02.2026 |
#14/23
better than 41%
|
|
MMStar
vision_language
|
65.52%
|
01.02.2026 |
#12/16
better than 27%
|
|
HallusionBench
vision_language
|
27.05%
|
01.02.2026 |
#12/13
better than 8%
|
|
OmniDocBench 1.5
document_understanding
|
84.03%
|
01.02.2026 |
#19/26
better than 28%
|
|
CharXiv (RQ)
document_understanding
|
10.28%
|
01.02.2026 |
#27/34
better than 21%
|
|
MMLongBench-Doc
document_understanding
|
50.30
|
01.02.2026 |
#7/7
better than 0%
|
|
CC-OCR
document_understanding
|
19.42%
|
01.02.2026 |
#13/14
better than 8%
|
|
OCRBench
document_understanding
|
76.60
|
01.02.2026 |
#5/6
better than 0%
|
|
CountBench
spatial_intelligence
|
12.82%
|
01.02.2026 |
#11/13
better than 17%
|
|
RefSpatialBench
spatial_intelligence
|
10.03%
|
01.02.2026 |
#10/12
better than 18%
|
|
LingoQA
spatial_intelligence
|
71.68%
|
01.02.2026 |
#7/8
better than 14%
|
|
TIR-Bench
vision_language
|
24.60
|
01.02.2026 |
#8/8
better than 0%
|
|
V-Star
vision_language
|
32.91%
|
01.02.2026 |
#17/20
better than 16%
|
|
SLAKE
vision_language
|
58.74%
|
01.02.2026 |
#6/8
better than 29%
|
|
PMC-VQA
vision_language
|
36.30
|
01.02.2026 |
#8/8
better than 0%
|
|
MedXpertQA-MM
vision_language
|
24.89%
|
01.02.2026 |
#11/13
better than 17%
|
|
HLE (with tools)
stem_reasoning
|
39.04%
|
01.02.2026 |
#40/49
better than 19%
|
|
OCRBench
document_understanding
|
0.75%
|
01.02.2026 |
#11/12
better than 9%
|
|
BrowseComp
general_agent
|
50.56%
|
01.02.2026 |
#29/45
better than 36%
|
|
C-Eval
knowledge
|
82.20
|
01.02.2026 |
#15/15
better than 0%
|
|
AI2D_TEST
document_understanding
|
83.06%
|
01.02.2026 |
#10/15
better than 36%
|
|
MMBench EN-DEV v1.1
vision_language
|
68.62%
|
01.02.2026 |
#13/15
better than 14%
|
|
MMVU
video_understanding
|
57.88%
|
01.02.2026 |
#8/10
better than 22%
|
|
VideoMME (w sub.)
video_understanding
|
69.05%
|
01.02.2026 |
#13/17
better than 19%
|
|
SimpleVQA
vision_language
|
54.96%
|
01.02.2026 |
#10/16
better than 40%
|
|
MLVU
video_understanding
|
75.00%
|
01.02.2026 |
#9/13
better than 33%
|
|
ERQA
spatial_intelligence
|
45.86%
|
01.02.2026 |
#11/21
better than 45%
|
|
IFBench
instruction_following
|
85.69%
|
01.02.2026 |
#18/42
better than 59%
|
|
DeepPlanning
general_agent
|
3.61%
|
01.02.2026 |
#7/10
better than 33%
|
|
VideoMMMU
video_understanding
|
85.90%
|
01.02.2026 |
#9/19
better than 56%
|
|
AA-LCR
long_context
|
80.95%
|
01.02.2026 |
#16/42
better than 63%
|
|
MathVista (mini)
vision_language
|
65.61%
|
01.02.2026 |
#13/14
better than 8%
|
|
VideoMME (w/o sub.)
video_understanding
|
80.92%
|
01.02.2026 |
#7/11
better than 40%
|
|
EmbSpatialBench
spatial_intelligence
|
72.24%
|
01.02.2026 |
#10/12
better than 18%
|
|
BabyVision
vision_language
|
10.18%
|
01.02.2026 |
#20/24
better than 17%
|
What is a good MMMU-Pro score? What is a good MMLU-Pro score? What is a good MMLU-Redux score?
No comments yet
Be the first to share your thoughts!