Model
GPT-Image-2
OpenAI · closed weights
Agentic / Tool Use
Agentic / Tool UseMultimodal
Multimodal| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| EmoGenie | 83.0 | accuracy | N/A | Independent | medium | arxiv.org |
| MTEdit-Bench | 6.6 | mtedit-bench_score | N/A | Independent | medium | arxiv.org |
| RISEBench | 56.8 | accuracy | N/A | Independent | medium | arxiv.org |
| UltraText Bench | 99.3 | composite | N/A | Independent | medium | arxiv.org |
| VeriEdit-Bench | 61.5 | attainment | N/A | Independent | medium | arxiv.org |
| VisSlideBench | 8.4 | overall | N/A | Independent | medium | arxiv.org |
Safety & Alignment
Safety & Alignment| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| EpiReal-Bench | 77.0 | success_rate | N/A | Independent | medium | arxiv.org |
Reproduced by the community
Loading community runs…