GPT 5.2
OpenAI
Highlights
Top benchmark results for openai/gpt-5-2-2025-12-11.
1#2
0.86#3
0.53#2
0.40#1
0.92#3
0.99#2
0.46#3
0.90#3
0.56#1
0.80#2
0.75#1
Benchmark table
| Benchmark | Category | Top Score | Info | Self Reported | Source |
|---|---|---|---|---|---|
| AIME 2025 | math | 1 | No Tools | Yes | Source |
| ARC-AGI-1 | - | 0.86 | Extra High Reasoning Effort | Yes | - |
| ARC-AGI-2 | - | 0.53 | Extra High Reasoning Effort | Yes | - |
| Frontier Math | math | 0.40 | Tier 1-3 With Python | Yes | Source |
| GPQA Diamond | general-knowledge | 0.92 | No Tools | Yes | Source |
| HMMT 2025 | - | 0.99 | Feb 2025, No Tools | Yes | Source |
| Humanity's Last Exam | - | 0.46 | With Search + Python | Yes | Source |
| MMMLU | - | 0.90 | - | Yes | Source |
| SWE Bench Pro | - | 0.56 | - | Yes | Source |
| SWE-Bench | code | 0.80 | - | Yes | Source |
| SWE-Lancer | code | 0.75 | IC Diamond | Yes | Source |
Benchmark comparisons
Use the selector to switch benchmarks and see how this model stacks up against its closest competitors.
SWE Bench Pro
Compare this model with the leading peers for the selected benchmark.
Benchmark
0.56
Rank #1/2
2 models
Showing 2 models around the selected model (out of 2 total).