EditJev: System One Routing for Image Editing Models

Zhile Guan; Xu Jiang; Yule Duan; Gongbo Zhang; Shuyu Wang; Xiao Chen; Jian Zhang

cs.CVOct 23, 2026

EditJev: System One Routing for Image Editing Models

Authors: Zhile Guan, Xu Jiang, Yule Duan, Gongbo Zhang, Shuyu Wang, Xiao Chen, Jian Zhang

Organizations: School of Electronic and Computer Engineering, Peking University · Tsinghua University

Abstract

Instruction-based image editing models exhibit pronounced sample-specific performance variation: different models can produce substantially different results for the same source image and editing instruction, and the best-performing model often varies across requests. These models also differ substantially in inference latency, making fixed-model deployment inefficient. We therefore study image editing model routing, which dynamically selects an editor from a candidate pool given an input image and editing instruction. We introduce EditJev, a System One routing model that directly predicts sample-level preference scores over candidate image editing models in a single forward pass. A lightweight preference head on a vision-language backbone produces a fixed-dimensional score vector without autoregressive model-name generation or response parsing. To train EditJev, we construct pairwise preferences from offline candidate outputs and optimize a pairwise ranking objective to learn relative model suitability. At inference time, EditJev combines the predicted preference scores with model-level speed scores to balance editing quality and generation latency. Experiments on ImgEdit and GEdit-En show that EditJev improves aggregate quality over the strongest fixed editor, from 4.540 to 4.592 and from 8.194 to 8.255, respectively, while reducing estimated end-to-end latency by 60.8% and 31.7%. These results demonstrate the effectiveness of direct, structured model routing for efficient image editing.

Figures & tables

Figure 1: Left: Different image editing models exhibit complementary strengths: the optimal model varies across image-instruction pairs, and no single model consistently performs best for all editing requests. Middle: This observation motivates model routing, where EditJev achieves a better quality–latency trade-off by dynamically selecting suitable models instead of relying on a fixed model. Right: Given an input image and instruction, EditJev uses a multimodal router to understand the request, estimate model suitability, and select the model that provides better editing quality with lower inference latency.

Figure 2: Overview of EditJev. The upper part shows offline candidate generation, automatic evaluation, and router training. The lower part shows quality–latency-aware inference, which combines preference scores with model-level speed scores.

Fixed Models
Method	Add	Adjust	Extract	Replace	Remove	Background	Style	Compose	Action	Overall $\uparrow$	Latency(s) $\downarrow$
FLUX.2-Klein-4B	4.470	4.660	1.690	4.480	4.300	4.470	4.850	3.510	4.800	4.073	–
FLUX.2-Klein-9B	4.580	4.730	1.950	4.680	4.710	4.680	4.910	3.590	4.880	4.252	–
FLUX.1-Kontext-dev	4.050	3.930	1.950	4.440	3.420	3.940	4.400	2.460	4.240	3.670	–
OmniGen2	3.730	3.640	1.640	4.040	3.110	3.680	4.580	2.670	4.420	3.461	–
Qwen-Image-Edit	4.660	4.650	4.120	4.750	4.630	4.410	4.770	3.770	4.750	4.540	–

Table 1: Quantitative comparison on ImgEdit. Methods are grouped into fixed models, MLLM-based routing methods, and System One routing methods. The best and second-best results for each editing category, overall quality, and latency are marked in bold and underline, respectively. Latency denotes the additional routing overhead before executing the selected editor.

Figure 3: Qualitative routing results on GEdit-En and ImgEdit. A red border marks the model selected by EditJev for each input.

Fixed Models
Method	G_SC	G_PQ	G_O $\uparrow$
FLUX.2-Klein-4B	7.731	8.397	7.667
FLUX.2-Klein-9B	8.081	8.416	7.965
FLUX.1 Kontext	6.404	7.579	6.241
OmniGen2	6.391	7.762	6.297
Qwen-Image-Edit	8.441	8.326	8.194

Table 2: Quantitative evaluation results on GEdit-En.

Method	ImgEdit (s) $\downarrow$	GEdit-En (s) $\downarrow$
Qwen-Image-Edit	56.90	56.90
GPT-5.5 Routing	51.15	53.55
EditJev	22.33	38.88

Table 3: Latency comparison among Qwen-Image-Edit, GPT-5.5 Routing, and EditJev. Latency includes both the routing inference time and the generation time of the selected editor.

Training Objective	Overall $\uparrow$
MSE Regression	4.5545
MCQ Generation	4.5491
Pairwise Ranking (Ours)	4.6002

Table 4: Ablation on the router training objective on the ImgEdit benchmark.

Variant	Overall $\uparrow$
w/o Image Input	4.5717
w/ Image Input (Ours)	4.6002

Table 5: Ablation on the visual input to the router on ImgEdit.

Router Backbone	Overall $\uparrow$
Qwen3-VL-2B	4.5898
Qwen3-VL-4B (Ours)	4.6002
Qwen3-VL-32B	4.5898

Table 6: Ablation on the router backbone scale on ImgEdit.

Figure 4: Quality–latency trade-off on ImgEdit under different speed weights.

Method	Agreement (%) $\uparrow$	Spearman $\rho$ $\uparrow$
GPT-5.5 Routing	34.50	—
EditJev (Ours)	81.75	0.7478

Table 7: Human preference alignment on 20 cases from ImgEdit and GEdit-En.

RewardHarness: Self-Evolving Agentic Post-Training
May 9, 2026Yuxuan Zhang, Cong Wei, Penghui Du +11Reward Modeling Image Editing Evaluation
University of British Columbia · Vector Institute · Kolors Team, Kuaishou Technology +5
From Scale to Speed: Adaptive Test-Time Scaling for Image Editing
Feb 24, 2026Xiangyan Qu, Zhenlong Yuan, Jing Tang +9Test-Time Scaling Text-Guided Image Editing
Institute of Information Engineering, Chinese Academy of Sciences · School of Cyber Security, University of Chinese Academy of Sciences · AMAP, Alibaba Group +1
HP-Edit: A Human-Preference Post-Training Framework for Image Editing
Apr 21, 2026Fan Li, Chonghuinan Wang, Lina Lei +9Diffusion Model Alignment Image Editing Evaluation
Huawei Noah’s Ark Lab · Nankai University · Harbin Institute of Technology

May 9, 2026J/K move · Enter open · S save

RewardHarness: Self-Evolving Agentic Post-Training

Yuxuan Zhang, Cong Wei, Penghui Du, Bo Li+10

University of British Columbia · Vector Institute · Kolors Team, Kuaishou Technology+5

Reward Modeling Image Editing Evaluation Multimodal Reward Modeling Image Generation Evaluation

Open paper AskPDF arXiv

Aligning generative models with human preferences remains a central challenge. While RLHF is widely adopted, it relies on large-scale preference data. This limitation is particularly evident in evaluating instruction-guided image edits, even though humans can often learn such preferences from only a few examples. We present RewardHarness, a training-free, self-evolving agentic reward framework that reframes reward modeling as context evolution rather than weight optimization. Instead of learning from large-scale annotations, RewardHarness aligns with human preferences by iteratively evolving a library of tools and skills from as few as 100 preference demonstrations. Given a source image, an edited image, and an editing instruction, an orchestrator agent selects the most relevant subset of tools and skills from the maintained library, and sub-agents invoke them to construct a reasoning chain that produces a reward score. By comparing predicted scores with ground-truth preferences and analyzing successes and failures in the reasoning process, the orchestrator automatically refines its library of tools and skills without additional human annotation. Using only 0.05% of the EditReward preference data, RewardHarness achieves 47.4% average accuracy on EditReward-Bench K=2/3/4 and GenAI-Bench, surpassing GPT-5 by 5.3 points. When used as a reward signal in FlowGRPO, RL-tuned models achieve 3.52 on ImgEdit-Bench. Code and model are available at the project page: https://rewardharness.com/

Abstract

Explore similar work

RewardHarness: Self-Evolving Agentic Post-Training

From Scale to Speed: Adaptive Test-Time Scaling for Image Editing

HP-Edit: A Human-Preference Post-Training Framework for Image Editing