- Humanity's Last Exam (Full Set), browser + computer + terminal 41.6
- FrontierMath Tier 1-3, with tools (text) 27.4
- DSBench: Data Analysis 89.9 (Data Modeling 85.5) 89.9
- SpreadsheetBench, OSX LibreOffice, with .xlsx access 45.5 (without 35.3) 45.5
- Investment Banking Modeling Tasks, mean accuracy (Oracle@64 71.3) 41
- BrowseComp (Agentic Browsing) 68.9
- WebArena (Agentic Browser Use) 65.4
ChatGPT agent
ChatGPT agent ranks first on 2 of the 10 benchmarks it has scores on.
21
Benchmarks with a score
2
Ranked first, out of five or more models
3
In the top three
OpenAI Jul 2025
ChatGPT agent
Closed weights Flagship
No Epoch Capabilities Index score yet
- WebArena 65.4% #3 /8
- Remote Labor Index 1.25% #15 /17
- First-Person Fairness evaluation 0.40% #10 /10
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| SimpleQA Hallucination Rate | 7.90% | #1 of 12 | Model card | |
| SimpleQA | 91.4% | #5 of 59 | Model card | |
| PersonQA | 94.3% | #1 of 9 | Model card | |
| Humanity's Last Exam | 41.6% | #57 of 147 | Model card |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| FrontierMath | 27.4% | #6 of 11 | Model card |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| DSBench | 89.9% | #1 of 1 | Model card |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| WebArena | 65.4% | #3 of 8 | Model card | |
| BrowseComp | 68.9% | #45 of 77 | Model card | |
| AgentIF-OneDay | 62.6% | #1 of 1 | Official leaderboard |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| SpreadsheetBench | 45.5% | #5 of 6 | Model card | |
| Remote Labor Index | 1.25% | #15 of 17 | Official leaderboard | |
| Investment Banking Modeling Tasks (OpenAI) | 41% | #4 of 4 | Model card |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Filtered, adversarial sample of production prompts | 96.9% | #1 of 3 | Model card | |
| First-Person Fairness evaluation | 0.40% | #10 of 10 | Model card | |
| Challenging prompts from red teamers with biosafety-relevant PhDs | 87.9% | #3 of 3 | Model card | |
| Proactive refusals of high-risk tasks | 89% | #2 of 2 | Model card | |
| User Confirmations During Computer Use | 91% | #2 of 2 | Model card | |
| Active data exfiltration, visual browser | 67% | #1 of 1 | Model card | |
| In-context data exfiltration, visual browser | 78% | #1 of 1 | Model card | |
| Irrelevant instructions, text-based web browser | 99.5% | #1 of 1 | Model card | |
| Irrelevant instructions, visual browser | 95% | #1 of 1 | Model card |
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.
Its system card
Every result in the longer documents the lab published with it. 32 of them aren't in the launch post.
ChatGPT Agent System Card 17 Jul 2025
Hallucination
- SimpleQA 91.4
- SimpleQA Hallucination Rate 7.9
- PersonQA 94.3
Safety
- Standard Refusal Evaluation
- OpenAI Production Benchmarks and safe-completion evals
- Image input safety evaluations
- User Confirmations During Computer Use 91
- Agentic harmful task refusals
- Proactive refusals of high-risk tasks 89
- Challenging prompts from red teamers with biosafety-relevant PhDs 87.9
- Filtered, adversarial sample of production prompts 96.9