List your company

Structured Image Editing on StructEditBench

64.64Math AccuracyNano Banana 2.0
7.200822.112937.02551.9371Oct 6, 2025

Evaluation results

MethodLinks
Nano Banana 2.02025.1064.6424.88—63.2816.74—72.5826.39—81.0728.12—71.1926.71—91.125.14—67.0525.59—
Ours2025.1054.7423.31—50.5815.33—60.1824.65—7326.33—75.0525.8—77.0823.19—55.9824.01—
GPT-Image2025.1051.4917.06—45.8212.56—50.7117.24—76.0316.55—67.6116.61—83.2614.35—52.216.64—
Seedream 4.02025.1051.0623.63—46.8316.54—51.7224.12—71.1326.93—69.2226.46—88.1924.75—52.8524.45—
Nano Banana2025.1050.4620.77—46.4214.93—52.9721.22—66.5622.92—69.1622.61—75.7119.75—51.5721.09—
Seedream 4.0Evaluator=Qwen2.5-VL-72B2025.1041.3123.6395.8743.1716.5479.9241.1824.1294.3359.9426.9393.8453.3926.4695.1878.5624.7593.8346.0224.4595.14
OursEvaluator=Qwen2.5-VL-72B2025.1041.123.3195.1440.4415.3375.2837.224.6596.2657.5826.3396.4752.4625.896.3869.7523.1994.3843.5624.0185.35
Qwen-Edit2025.1040.4823.73—30.1712.33—44.8326.11—53.7427.31—55.9925.53—67.7625.71—38.1224.81—
GPT-ImageEvaluator=Qwen2.5-VL-72B2025.1039.7717.0691.9939.6112.5665.847.7717.2493.5759.0116.5591.3953.2316.6192.9378.0214.35844416.6491.19
Nano BananaEvaluator=Qwen2.5-VL-72B2025.1039.4620.7795.2642.3314.9373.9246.7221.2296.1654.3522.9296.2153.3322.6196.4666.8719.7592.644.4621.0995.24
Bagel-Think2025.1037.423.97—28.9816.82—42.5126.49—36.1126.75—43.1525.57—40.4623.83—33.3424.7—
FLUX.1 Kontext2025.1037.3619.78—32.2914.61—39.1220.1—58.3520.38—50.3920.99—58.0518.52—37.5619.84—
Step1X-Edit2025.1034.4723.41—28.0516.68—33.2624.56—60.4825.94—46.4724.98—57.8123.97—34.1124.03—
Qwen-EditEvaluator=Qwen2.5-VL-72B2025.1034.0323.7397.1425.512.3365.9338.2326.1198.1443.2527.3197.9443.0825.5397.4258.8225.7196.6531.9924.8197.31
Ovis-U12025.1031.6418.45—21.9413.3—38.0319.01—42.0817.92—44.5218.68—35.5816.62—28.0618.25—
Flux KontextEvaluator=Qwen2.5-VL-72B2025.1029.7819.7895.1227.7814.6173.2430.2520.195.5446.3220.3895.0135.620.9996.2752.2218.5291.2731.1319.8494.78
OmniGen22025.1029.4415.95—18.5512.44—34.6311.31—28.6116.51—39.5515.6—30.3616.6—24.315.49—
HiDream-E1.12025.1028.0718.43—26.3612.91—29.6318.26—43.7718.04—36.6616.47—48.7917.12—29.6318.01—
Bagel-ThinkEvaluator=Qwen2.5-VL-72B2025.1027.8223.9797.3423.3816.8282.2732.326.4998.2323.2226.7598.0933.1925.5798.0435.7323.8397.3126.1724.797.59
Step1X-EditEvaluator=Qwen2.5-VL-72B2025.1027.2223.4197.1124.9916.6881.5425.1124.5697.7543.9525.9497.8633.4924.9897.3848.4823.9796.8728.2624.0397.26
DiMOO2025.1026.7921.56—16.5214.98—24.0321.77—29.5222.26—26.0822.57—24.6419.47—2121.49—
Ovis-U1Evaluator=Qwen2.5-VL-72B2025.1025.5718.4593.5613.4813.368.5428.4219.0195.0124.5917.9292.42218.6894.5533.0316.6287.9619.418.2593
OmniGen2Evaluator=Qwen2.5-VL-72B2025.1023.6515.9582.0613.4412.4471.8723.6811.3178.3416.5616.5186.5426.0215.686.1527.5216.688.8818.115.4983.34
DiMOOEvaluator=Qwen2.5-VL-72B2025.1022.2221.5695.198.7814.9874.3716.321.779621.1822.2695.6818.1122.5796.2620.619.4792.0914.2221.4995.04
Hidream-E1.1Evaluator=Qwen2.5-VL-72B2025.1021.9718.4393.9920.2512.9170.5922.418.2694.6629.1618.0493.8925.4216.4793.7746.8317.1291.1823.0718.0193.68
Bagel2025.1021.2721.38—27.1116.38—29.9422.7—41.5924.22—47.1623.56—47.3521.54—28.8722.06—
BagelEvaluator=Qwen2.5-VL-72B2025.1016.0721.3896.0324.2716.3881.7820.3222.797.2735.4424.2297.2832.5823.5697.3645.3721.5495.4224.2422.0696.36
UniworldEvaluator=Qwen2.5-VL-72B2025.1015.978.8482.646.197.8748.8310.276.1672.729.987.7176.1916.017.6776.7215.088.2472.979.988.2178.95
UniWorld-V12025.109.418.84—5.997.87—8.836.16—9.117.71—19.917.67—16.138.24—8.48.21—