Computer-Using Agent (OpenAI) Jan 23, 2025
OpenAI Computer-Using Agent (CUA)
OpenAI Computer-Using Agent (CUA) ranks first on 2 of the 9 benchmarks with five or more models scored.
16
Benchmarks with a score
2
Ranked first, out of five or more models
3
In the top three
OpenAI Jan 2025
OpenAI Computer-Using Agent (CUA)
Closed weights Flagship
No Epoch Capabilities Index score yet
- WebArena 58.1% #6 /8
- WebVoyager 87% #5 /12
- StrongREJECT 66% #18 /19
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.
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| OSWorld-G | 40.6% | #1 of 3 | Official leaderboard | |
| WebVoyager | 87% | #5 of 12 | Lab model card | |
| OSWorld | 38.1% | #6 of 13 | Lab model card | |
| WebArena | 58.1% | #6 of 8 | Lab model card | |
| WebBench (Halluminate) | 59.8% | #1 of 1 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Human Sourced Jailbreaks | 100% | #1 of 8 | Lab model card | |
| Production Jailbreaks | 100% | #1 of 5 | Lab model card | |
| Standard Refusal Evaluation | 100% | #2 of 8 | Lab model card | |
| Challenging Refusal Evaluation | 92% | #4 of 9 | Lab model card | |
| Proactive refusals of high-risk tasks | 94% | #1 of 2 | Lab model card | |
| User Confirmations During Computer Use | 92% | #1 of 2 | Lab model card | |
| Jailbreak Augmented Examples | 100% | #4 of 5 | Lab model card | |
| StrongREJECT | 66% | #18 of 19 | Lab model card | |
| AutoElicit | 48.7% | #4 of 4 | Official leaderboard | |
| Agentic harmful task refusals | 97% | #1 of 1 | Lab model card | |
| Operator prompt injection scenarios | 23% | #1 of 1 | Lab 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. 12 of them aren't in the launch post.
Operator System Card Jan 23, 2025
Capabilities
- OSWorld 38.1