OpenAI
gpt-5-2025-08-07
proprietaryMultimodal
Benchmark Average (0–100)
68.9%
Average of non-null benchmark scores across all evaluated tasks.
BENCHMARK SCORES
GPQA85.7%
Graduate-Level Google-Proof Q&A. PhD-level multiple-choice questions in chemistry, biology, and physics. Scored by accuracy.
AIME 202594.6%
AIME 2025 is a mathematics competition benchmark testing advanced problem-solving.
SWE-bench Verified74.9%
SWE-bench Verified evaluates models on real-world software engineering tasks from GitHub issues.
BrowseComp54.9%
BrowseComp tests web browsing comprehension, measuring ability to find and synthesize information from websites.
HLE24.8%
Humanity's Last Exam (HLE) is a multi-modal benchmark testing frontier knowledge across mathematics, humanities, and natural sciences with 2,500 expert-level questions.
MMMU-Pro78.4%
MMMU-Pro is a more challenging version of MMMU with harder multimodal understanding tasks.
OTHER BENCHMARKS
MMMU
84.2%
CharXIV-R
81.1%
FrontierMath
26.3%
MODEL INFO
Organization
OpenAI
Context Window
400,000
Release Date
2025-08-07
Knowledge Cutoff
2024-09-30
Parameters
—
License
proprietary
Input Price ($/M)
$1.2500
Output Price ($/M)
$10.0000
METADATA
Announcement Date
2025-08-07
Organization Country
US
Canonical ID
—