List your company

Text-to-Image Generation on T2I-ReasonBench

75.7Idiom AccuracyGPT-4o
9.34826.57443.861.026Nov 14, 2025Dec 14, 2025Jan 14, 2026Feb 14, 2026Mar 17, 2026Apr 17, 2026May 18, 2026

Evaluation results

MethodLinks
GPT-4oType=proprietary2026.0175.794.586.997.677.596.674.794.378.795.8
gpt-image-1Types=Gen2025.1175.794.586.997.677.596.674.794.378.795.8
GPT-4oType=Proprietary2026.0575.794.586.997.677.596.674.794.378.795.8
Generation NavigatorType=Agent2026.0574.9694.7989.7194.6273.997.0877.6692.8379.0694.83
StruVisBase Model=Qwen3-VL-8B, Generator=Flux-dev-12026.0370.1592.9975.2284.7576.7497.3372.1892.1273.5791.8
StruVisType=Agent2026.0570.1592.9975.2284.7576.7497.3372.1892.1273.5791.8
StruVisBase Model=Qwen2.5-VL-7B, Generator=Flux-dev-12026.0369.8593.0874.6484.7576.396.8372.0993.2973.2291.99
InterleavedBase Model=Qwen3-VL-8B, Generator=Flux-dev-12026.0361.9292.3174.9284.7170.5695.0468.4691.7568.9790.95
InterleavedBase Model=Qwen2.5-VL-7B, Generator=Flux-dev-12026.0361.4691.7972.984.5463.5595.9867.4891.0866.3590.85
Text-OnlyBase Model=Qwen3-VL-8B, Generator=Flux-dev-12026.0361.4391.0874.0382.9662.0394.4862.488.2564.9789.19
W/O.Base Model=Qwen3-VL-8B, Generator=Flux-dev-12026.0360.6690.7771.3582.7159.2594.1361.7790.4263.2689.51
Ours (w/o GRPO)Type=diffusion2026.0159.590.471.587.161.2993.972.291.866.190.8
OursType=diffusion2026.0158.590.675.289.568.895.272.993.568.392.2
Think-then-GenerateType=Agent2026.0558.590.675.289.568.895.272.993.568.392.2
FLUX.2-Klein-9BType=Open-source2026.0555.3593.8343.2177.0257.7884.2974.4476.1257.782.82
Gemini-2.0Type=proprietary2026.0152.487.87383.36794.366.789.364.888.7
Gemini-2.0Types=Unified2025.1152.487.87383.36794.366.789.364.888.7
Gemini-2.0Type=Proprietary2026.0552.487.87383.36794.366.789.364.888.7
Text-OnlyBase Model=Qwen2.5-VL-7B, Generator=Flux-dev-12026.0351.7190.3172.984.5461.1695.963.8789.7562.4190.12
Ours (w/o SFT+GRPO)Type=diffusion2026.0151.786.371.783.857.392.862.887.660.986.6
UniCoTType=unified2026.014984.258.192.373.592.951.971.758.185.3
UniCoTType=Open-source2026.054984.258.192.373.592.951.971.758.185.3
Qwen-ImageType=diffusion2026.0148.184.366.585.857.184.759.585.357.887.5
Qwen-ImageType=Open-source2026.0548.184.366.585.857.184.759.585.357.887.5
W/O.Base Model=Qwen2.5-VL-7B, Generator=Flux-dev-12026.0346.3290.5864.9578.559.9694.4858.7188.3357.4887.97
Bagel w/CoTType=unified2026.0144.684.34473.752.491.657.788.349.784.5
Self-RewriteTypes=Bagel, Backbone=Bagel2025.1144.684.34473.752.491.657.788.349.784.5
Bagel w/CoTType=Open-source2026.0544.684.34473.752.491.657.788.349.784.5
playground-v2.5Types=Gen2025.1143.987.838.572.148.492.450.883.345.483.9
FLUX.1-schnellTypes=Gen2025.1140.983.165.174.544.891.550.78350.483
FLUX.1-devType=diffusion2026.0139.183.456.976.545.190.646.780.94782.8
FLUX.1-devTypes=Gen2025.1139.183.456.976.545.190.646.780.94782.8
ImAgentTypes=Bagel, Backbone=Bagel2025.1137.79054.279.152.696.661.290.351.489
SD-3-MediumTypes=Gen2025.1135.981.460.971.342.490.150.981.747.581.1
SD-3.5-largeType=diffusion2026.0135.685.362.275.446.692.652.984.549.384.4
SD-3.5-LargeTypes=Gen2025.1135.685.362.275.446.692.652.984.549.384.4
SD-3.5-largeType=Open-source2026.0535.685.362.275.446.692.652.984.549.384.4
SD-3.5-mediumType=diffusion2026.0134.480.65870.144.892.149.98346.881.4
SD-3.5-MediumTypes=Gen2025.1134.480.65870.144.892.149.98346.881.4
Emu3Types=Unified2025.1133.182.933.768.733.885.240.17735.278.5
show-oTypes=Unified2025.1133.182.535.380.334.987.441.676.636.281.7
VanillaTypes=Bagel, Backbone=Bagel2025.1130.285.736.668.44594.754.487.541.684.1
GoTTypes=Unified2025.1129.776.430.670.73186.236.876.33277.4
ImAgentTypes=Janus-Pro-7B, Backbone=Janus-Pro-7B2025.1127.98635.368.840.789.751.284.238.882.2
Janus-Pro-7BType=unified2026.0125.57837.270.938.587.644.977.836.578.6
VanillaTypes=Janus-Pro-7B, Backbone=Janus-Pro-7B2025.1125.57837.270.938.587.644.977.836.578.6
HunyuanImage-3.0Type=unified2026.0125.480.254.280.952.392.256.884.447.284.4
HunyuanImage-3.0Type=Open-source2026.0525.480.254.280.952.392.256.884.447.284.4
Self-RewriteTypes=Janus-Pro-7B, Backbone=Janus-Pro-7B2025.1111.974.69.249.31274.31459.311.163.3