| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Multi-image reasoning | OmniContext | Single Scene Char Score8.96 | 20 | |
| In-context image generation | OmniContext 1.0 (test) | Single Instance Character Fidelity8.9 | 13 | |
| Multi-reference image generation | OmniContext SCENE 1.0 (test) | Character Fidelity Score8.9 | 10 | |
| Multi-reference image generation | OmniContext MULTIPLE 1.0 (test) | Character Score9.07 | 10 | |
| Subject-Driven Image Generation | OmniContext SINGLE 1.0 | Character PF Score8.89 | 10 | |
| In-context Image Generation | OmniContext (test) | Prompt Following7.58 | 5 | |
| In-context Video Generation | OmniContext-Video | Single ID Char Score6.71 | 3 |