Interleaved Image-Text Generation on RISEBench
34.1Temporal CoherenceGPT-4o
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GPT-4oType=Private, Evaluator=GPT-4.12025.05 | 34.1 | 32.2 | 37 | 10.6 | 28.9 | |
| GPT-4oTypes=Private2025.11 | 34.1 | 32.2 | 37 | 10.6 | 28.9 | |
| ImAgentTypes=Bagel2025.11 | 17.6 | 15.6 | 16 | 2.4 | 13.1 | |
| Seedream-4.0Types=Open2025.11 | 12.9 | 12.2 | 11 | 7.1 | 10.8 | |
| Gemini-2.0-Flash-preTypes=Private2025.11 | 10.6 | 13.3 | 11 | 2.3 | 9.4 | |
| Gemini 2.0Type=Private, Evaluator=GPT-4.12025.05 | 8.2 | 15.5 | 23 | 4.7 | 13.3 | |
| Gemini-2.0Types=Private2025.11 | 8.2 | 15.5 | 23 | 4.7 | 13.3 | |
| BAGELType=Open-source, Evaluator=GPT-4.1, Self-CoT=true2025.05 | 5.9 | 17.8 | 21 | 1.2 | 11.9 | |
| Self-RewriteTypes=Bagel2025.11 | 5.9 | 17.8 | 21 | 1.2 | 11.9 | |
| Qwen-Image-EditTypes=Open2025.11 | 4.7 | 10 | 17 | 2.4 | 8.9 | |
| BAGELType=Open-source, Evaluator=GPT-4.1, Self-CoT=false2025.05 | 2.4 | 5.6 | 14 | 1.2 | 6.1 | |
| VanillaTypes=Bagel2025.11 | 2.4 | 5.6 | 14 | 1.2 | 6.1 | |
| FLUX.1-Kontext-DevTypes=Open2025.11 | 2.3 | 5.5 | 13 | 1.2 | 5.8 | |
| EMU2Type=Open-source, Evaluator=GPT-4.12025.05 | 1.2 | 1.1 | 0 | 0 | 0.5 | |
| OmniGenType=Open-source, Evaluator=GPT-4.12025.05 | 1.2 | 1 | 0 | 1.2 | 0.8 | |
| EMU2Types=Open2025.11 | 1.2 | 1.1 | 0 | 0 | 0.5 | |
| OmniGenTypes=Open2025.11 | 1.2 | 1 | 0 | 1.2 | 0.8 | |
| Ovis-U1Types=Open2025.11 | 1.2 | 3.3 | 4 | 2.4 | 2.8 | |
| Step1X-EditType=Open-source, Evaluator=GPT-4.12025.05 | 0 | 2.2 | 2 | 3.5 | 1.9 | |
| Step1X-EditTypes=Open2025.11 | 0 | 2.2 | 2 | 3.5 | 1.9 |