# Text-to-Image on UniGenBench Short Source: [Text-to-Image on UniGenBench Short]() Collection index: [llms.txt]() Data updated: 2026-06-02T07:22:53.562Z Evaluation of Text-to-Image models on the UniGenBench Short dataset, assessed using the UniGen Score. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Text-to-Image | | Dataset (including any stated split) | UniGenBench Short | | Metric | UniGen Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 29 of 29 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [UniGen Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | UniGen Score | Result date | Report paper | | --- | --- | --- | --- | | Gemini-3.1-Pro (MSP) | 86.69 | Not recorded | 6a1e636edbd7150506f850e0 | | Gemini-3.1-Pro (MSP) | 85.53 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE | 85.39 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE | 84.6 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE | 83.56 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE | 83.34 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE | 82.68 | Not recorded | 6a1e636edbd7150506f850e0 | | Gemini-3.1-Pro (MSP) | 82.63 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE - RL | 82.33 | Not recorded | 6a1e636edbd7150506f850e0 | | No Enhancer | 80.58 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE | 80.42 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-4B (MSP) | 77.21 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE | 77.16 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE | 77.1 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-1.7B | 76.45 | Not recorded | 6a1e636edbd7150506f850e0 | | MAPE - RL | 76.23 | Not recorded | 6a1e636edbd7150506f850e0 | | Gemini-3.1-Pro (MSP) | 75.01 | Not recorded | 6a1e636edbd7150506f850e0 | | No Enhancer | 74.93 | Not recorded | 6a1e636edbd7150506f850e0 | | No Enhancer | 74.89 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-1.7B (MSP) | 73.83 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-4B | 72.21 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-1.7B | 70.93 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-4B | 70.55 | Not recorded | 6a1e636edbd7150506f850e0 | | Gemini-3.1-Pro | 69.49 | Not recorded | 6a1e636edbd7150506f850e0 | | No Enhancer | 69.31 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-1.7B | 68.78 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-1.7B | 67.9 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-4B | 48.17 | Not recorded | 6a1e636edbd7150506f850e0 | | Qwen3-4B | 47.77 | Not recorded | 6a1e636edbd7150506f850e0 | ### Gemini-3.1-Pro (MSP) UniGen Score: 86.69. Result ID: 6a1e84cafbc373592c1839f1. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"Qwen-Image-2512"}. This record does not assert independent reproduction. ### Gemini-3.1-Pro (MSP) UniGen Score: 85.53. Result ID: 6a1e84cafbc373592c183a08. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-9B"}. This record does not assert independent reproduction. ### MAPE UniGen Score: 85.39. Result ID: 6a1e84cafbc373592c1839f5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Qwen-Image-2512","Base Enhancer":"Qwen3-4B"}. This record does not assert independent reproduction. ### MAPE UniGen Score: 84.6. Result ID: 6a1e84cafbc373592c183a0c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-9B","Base Enhancer":"Qwen3-4B"}. This record does not assert independent reproduction. ### MAPE UniGen Score: 83.56. Result ID: 6a1e84cafbc373592c183a00. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo","Base Enhancer":"Qwen3-4B"}. This record does not assert independent reproduction. ### MAPE UniGen Score: 83.34. Result ID: 6a1e84cafbc373592c1839f3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Qwen-Image-2512","Base Enhancer":"Qwen3-1.7B"}. This record does not assert independent reproduction. ### MAPE UniGen Score: 82.68. Result ID: 6a1e84cafbc373592c183a0a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-9B","Base Enhancer":"Qwen3-1.7B"}. This record does not assert independent reproduction. ### Gemini-3.1-Pro (MSP) UniGen Score: 82.63. Result ID: 6a1e84cafbc373592c183a02. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-4B"}. This record does not assert independent reproduction. ### MAPE - RL UniGen Score: 82.33. Result ID: 6a1e84cafbc373592c1839ff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo","Base Enhancer":"Qwen3-4B"}. This record does not assert independent reproduction. ### No Enhancer UniGen Score: 80.58. Result ID: 6a1e84cafbc373592c183a07. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-9B"}. This record does not assert independent reproduction. ### MAPE UniGen Score: 80.42. Result ID: 6a1e84cafbc373592c183a06. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-4B","Base Enhancer":"Qwen3-4B"}. This record does not assert independent reproduction. ### Qwen3-4B (MSP) UniGen Score: 77.21. Result ID: 6a1e84cafbc373592c1839fe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo"}. This record does not assert independent reproduction. ### MAPE UniGen Score: 77.16. Result ID: 6a1e84cafbc373592c1839fc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo","Base Enhancer":"Qwen3-1.7B"}. This record does not assert independent reproduction. ### MAPE UniGen Score: 77.1. Result ID: 6a1e84cafbc373592c183a04. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-4B","Base Enhancer":"Qwen3-1.7B"}. This record does not assert independent reproduction. ### Qwen3-1.7B UniGen Score: 76.45. Result ID: 6a1e84cafbc373592c1839f2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Qwen-Image-2512"}. This record does not assert independent reproduction. ### MAPE - RL UniGen Score: 76.23. Result ID: 6a1e84cafbc373592c1839fb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo","Base Enhancer":"Qwen3-1.7B"}. This record does not assert independent reproduction. ### Gemini-3.1-Pro (MSP) UniGen Score: 75.01. Result ID: 6a1e84cafbc373592c1839f8. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo"}. This record does not assert independent reproduction. ### No Enhancer UniGen Score: 74.93. Result ID: 6a1e84cafbc373592c1839f0. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"Qwen-Image-2512"}. This record does not assert independent reproduction. ### No Enhancer UniGen Score: 74.89. Result ID: 6a1e84cafbc373592c183a01. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-4B"}. This record does not assert independent reproduction. ### Qwen3-1.7B (MSP) UniGen Score: 73.83. Result ID: 6a1e84cafbc373592c1839fa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo"}. This record does not assert independent reproduction. ### Qwen3-4B UniGen Score: 72.21. Result ID: 6a1e84cafbc373592c1839fd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo"}. This record does not assert independent reproduction. ### Qwen3-1.7B UniGen Score: 70.93. Result ID: 6a1e84cafbc373592c183a09. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-9B"}. This record does not assert independent reproduction. ### Qwen3-4B UniGen Score: 70.55. Result ID: 6a1e84cafbc373592c1839f4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Qwen-Image-2512"}. This record does not assert independent reproduction. ### Gemini-3.1-Pro UniGen Score: 69.49. Result ID: 6a1e84cafbc373592c1839f7. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo"}. This record does not assert independent reproduction. ### No Enhancer UniGen Score: 69.31. Result ID: 6a1e84cafbc373592c1839f6. [Reporting paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo"}. This record does not assert independent reproduction. ### Qwen3-1.7B UniGen Score: 68.78. Result ID: 6a1e84cafbc373592c1839f9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"Z-Image-turbo"}. This record does not assert independent reproduction. ### Qwen3-1.7B UniGen Score: 67.9. Result ID: 6a1e84cafbc373592c183a03. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-4B"}. This record does not assert independent reproduction. ### Qwen3-4B UniGen Score: 48.17. Result ID: 6a1e84cafbc373592c183a05. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-4B"}. This record does not assert independent reproduction. ### Qwen3-4B UniGen Score: 47.77. Result ID: 6a1e84cafbc373592c183a0b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"T2I Model":"FLUX.2-klein-9B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.