| KV-Edit: Training-Free Image Editing for Precise Background Preservation | ✓ Link | 35.87 | 9.92 | 1.98 | 25.63 | | KV-Edit | 2025-02-24 |
| FlashIn: Fast and Accurate Image Inversion for Real-time Image Editing | | 31.91 | | | 25.67 | 23.94 | FlashIn | 2026-06-03 |
| Multi-History-Step SDE Inversion for Image Editing with Superior Regional Awareness | | 29.35 | | | | | MIEdit (SD3.5) | 2026-09-08 |
| Inversion-Free Image Editing with Natural Language | ✓ Link | 28.51 | 47.58 | 13.78 | 25.03 | | Virtual Inversion+Unified Attention Control+LCM | 2023-12-07 |
| Vision-guided and Mask-enhanced Adaptive Denoising for Prompt-based Image Editing | ✓ Link | 28.27 | 45.67 | 12.65 | 25.91 | | Virtual Inversion+ViMAEdit | 2024-10-14 |
| Analyzing and Improving Training-Free Fast Sampling of Text-to-Image Diffusion Models | ✓ Link | 27.59 | 64.87 | 52.46 | 25.76 | 22.7 | TORS (Flux.1-Kontext) | 2026-09-08 |
| Inversion-Free Image Editing with Natural Language | ✓ Link | 27.52 | 47.98 | 14.22 | 24.89 | | Virtual Inversion+Prompt-to-Prompt | 2023-12-07 |
| Direct Inversion: Boosting Diffusion-based Editing with 3 Lines of Code | ✓ Link | 27.22 | 54.55 | 11.65 | 25.02 | | Direct Inversion+Prompt-to-Prompt | 2023-10-02 |
| Null-text Inversion for Editing Real Images using Guided Diffusion Models | ✓ Link | 27.03 | 60.67 | 13.44 | 24.75 | | Null-Text Inversion+Prompt-to-Prompt | 2022-11-17 |
| Target-aware Image Editing via Cycle-consistent Constraints | ✓ Link | 26.83 | | | 25.48 | | FlowCycle | 2025-10-23 |
| Inversion-Free Image Editing with Natural Language | ✓ Link | 26.64 | 55.85 | 15.61 | 24.57 | | Virtual Inversion+Prompt-to-Prompt+LCM | 2023-12-07 |
| Negative-prompt Inversion: Fast Image Inversion for Editing with Text-guided Diffusion Models | | 26.21 | 69.01 | 16.17 | 24.61 | | Negative-Prompt Inversion+Prompt-to-Prompt | 2023-05-26 |
| Wavelet-Guided Semantic Signal Compensation for Inversion-Free Image Editing | ✓ Link | 26.17 | 54.02 | 30.95 | | 25.52 | WSSC | 2026-07-02 |
| StyleDiffusion: Prompt-Embedding Inversion for Text-Based Editing | ✓ Link | 26.05 | 66.10 | 11.65 | 24.78 | | StyleDiffusion+Prompt-to-Prompt | 2023-03-28 |
| TDEdit: A Unified Diffusion Framework for Text-Drag Guided Image Manipulation | | 25.66 | 68.77 | 15.09 | 24.81 | | TDEdit | 2025-09-26 |
| FlashEdit: Decoupling Speed, Structure, and Semantics for Precise Image Editing | ✓ Link | 25.33 | 62.25 | | 25.51 | 22.29 | FlashEdit | 2025-09-26 |
| The Promise of RL for Autoregressive Image Editing | ✓ Link | 23.71 | 112.4 | 35.0 | 24.44 | 21.34 | EARL SFT (S) → RL (S+C) | 2025-08-01 |
| FireFlow: Fast Inversion of Rectified Flow for Image Semantic Editing | ✓ Link | 23.03 | 123.6 | 27.1 | 26.02 | | FireFlow | 2024-12-10 |
| Direct Inversion: Boosting Diffusion-based Editing with 3 Lines of Code | ✓ Link | 22.64 | 87.94 | 24.70 | 24.38 | | Direct Inversion+MasaCtrl | 2023-10-02 |
| Direct Inversion: Boosting Diffusion-based Editing with 3 Lines of Code | ✓ Link | 22.46 | 106.06 | 24.29 | 25.41 | | Direct Inversion+Plug-and-Play | 2023-10-02 |
| Plug-and-Play Diffusion Features for Text-Driven Image-to-Image Translation | ✓ Link | 22.28 | 113.46 | 28.22 | 25.41 | | DDIM Inversion+Plug-and-Play | 2022-11-22 |
| MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing | ✓ Link | 22.17 | 106.62 | 28.38 | 23.96 | | DDIM Inversion+MasaCtrl | 2023-04-17 |
| Direct Inversion: Boosting Diffusion-based Editing with 3 Lines of Code | ✓ Link | 21.53 | 138.98 | 49.22 | 23.31 | | Direct Inversion+Pix2Pix-Zero | 2023-10-02 |
| Zero-shot Image-to-Image Translation | ✓ Link | 20.44 | 172.22 | 61.68 | 22.80 | | DDIM Inversion+Pix2Pix-Zero | 2023-02-06 |
| Prompt-to-Prompt Image Editing with Cross Attention Control | ✓ Link | 17.87 | 208.80 | 69.43 | 25.01 | | DDIM Inversion+Prompt-to-Prompt | 2022-08-02 |
| FireFlow: Fast Inversion of Rectified Flow for Image Semantic Editing | ✓ Link | 16.49 | 239.4 | 70.9 | 27.33 | | FireFlow (Add Q) | 2024-12-10 |
| Visual Autoregressive Modeling for Instruction-Guided Image Editing | ✓ Link | | | | 26.0 | 22.5 | VAREdit 2.2B (512px) | 2025-08-21 |
| Source-Agnostic Image Translation Based on Latent Aware Adaptive Masking | ✓ Link | | | 19.68 | 26.43 | 22.81 | PM-Edit | 2026-08-14 |