GPT 5.5
OpenAI
Parameters
—
Architecture
—
Released
—
License
—
About
No detailed description available for this model.
Benchmark Scores
| Benchmark | Score | Date | Standing |
|---|---|---|---|
|
SWE-bench Verified
coding_agent
|
94.61%
|
11.06.2026 |
#2/60
better than 98%
|
|
Terminal Bench 2.1
coding_agent
|
86.25%
|
11.06.2026 |
#19/44
better than 58%
|
|
SWE Atlas - QnA
coding_agent
|
66.06%
|
11.06.2026 |
#9/24
better than 65%
|
|
SWE Atlas - TW
coding_agent
|
59.41%
|
11.06.2026 |
#7/20
better than 68%
|
|
SWE-efficiency
coding_agent
|
100.00%
|
11.06.2026 |
#1/5
better than 100%
|
|
LiveSQLBench
coding_agent
|
89.74%
|
11.06.2026 |
#2/5
better than 50%
|
|
CL-bench
coding_agent
|
100.00%
|
11.06.2026 |
#1/5
better than 100%
|
|
VIBE-V2
coding_agent
|
80.65%
|
11.06.2026 |
#2/9
better than 88%
|
|
SVG-Bench
coding_agent
|
63.35%
|
11.06.2026 |
#7/9
better than 25%
|
|
PostTrainBench
coding_agent
|
89.42%
|
11.06.2026 |
#4/22
better than 81%
|
|
KernelBench Hard
coding_agent
|
51.49%
|
11.06.2026 |
#3/5
better than 50%
|
|
PaperBench
coding_agent
|
43.11%
|
11.06.2026 |
#7/10
better than 33%
|
|
GDPVal
general_agent
|
91.83%
|
11.06.2026 |
#2/20
better than 95%
|
|
BankerToolBench
general_agent
|
33.89%
|
11.06.2026 |
#12/15
better than 21%
|
|
SpreadSheetBench-v1
general_agent
|
97.62%
|
11.06.2026 |
#3/9
better than 75%
|
|
YC-Bench
general_agent
|
59.09%
|
11.06.2026 |
#3/8
better than 71%
|
|
MCP-Atlas
general_agent
|
81.85%
|
11.06.2026 |
#16/54
better than 70%
|
|
OSWorld-Verified
general_agent
|
87.60%
|
11.06.2026 |
#5/25
better than 83%
|
|
OmniDocBench 1.5
document_understanding
|
95.52%
|
11.06.2026 |
#11/26
better than 60%
|
|
OfficeQA Pro
general_agent
|
69.98%
|
11.06.2026 |
#11/14
better than 23%
|
|
MMMU-Pro
vision_language
|
94.97%
|
11.06.2026 |
#3/35
better than 91%
|
|
Apex-Agents
general_agent
|
56.23%
|
11.06.2026 |
#4/24
better than 87%
|
|
BrowseComp
general_agent
|
91.35%
|
11.06.2026 |
#5/45
better than 91%
|
|
USAMO 2026
stem_reasoning
|
100.00%
|
11.06.2026 |
#1/4
better than 100%
|
|
IMO 2025
stem_reasoning
|
89.03%
|
11.06.2026 |
#2/4
better than 67%
|
|
NL2Repo-Bench
coding_agent
|
58.29%
|
11.06.2026 |
#12/23
better than 50%
|
|
VideoMME (w sub.)
video_understanding
|
97.14%
|
11.06.2026 |
#2/17
better than 94%
|
|
SWE-bench Pro
coding_agent
|
65.33%
|
11.06.2026 |
#20/77
better than 72%
|
|
VideoMMMU
video_understanding
|
96.08%
|
11.06.2026 |
#2/19
better than 94%
|
What is a good SWE-bench Verified score? What is a good Terminal Bench 2.1 score? What is a good SWE Atlas - QnA score?
No comments yet
Be the first to share your thoughts!