# Generative Fluid Intelligence on GENIUS Source: [Generative Fluid Intelligence on GENIUS]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:54:36.749Z Evaluation of model performance on the Generative Fluid Intelligence task using the GENIUS dataset, covering overall scores and detailed breakdowns across Implicit Pattern, Symbolic Constraint, Visual Constraint, Prior-Conflicting, and Multi-Semantic categories. 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 | Generative Fluid Intelligence | | Dataset (including any stated split) | GENIUS | | Metric | Overall Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 13 of 13 | | 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 - [Overall Score](): Higher is better; unit: PERCENT. - [Implicit Pattern RC](): Higher is better; unit: PERCENT. - [Implicit Pattern VC](): Higher is better; unit: PERCENT. - [Implicit Pattern AQ](): Higher is better; unit: PERCENT. - [Symbolic Constraint RC](): Higher is better; unit: PERCENT. - [Symbolic Constraint VC](): Higher is better; unit: PERCENT. - [Symbolic Constraint AQ](): Higher is better; unit: PERCENT. - [Visual Constraint RC](): Higher is better; unit: PERCENT. - [Visual Constraint VC](): Higher is better; unit: PERCENT. - [Visual Constraint AQ](): Higher is better; unit: PERCENT. - [Prior-Conflicting RC](): Higher is better; unit: PERCENT. - [Prior-Conflicting VC](): Higher is better; unit: PERCENT. - [Prior-Conflicting AQ](): Higher is better; unit: PERCENT. - [Multi-Semantic RC](): Higher is better; unit: PERCENT. - [Multi-Semantic VC](): Higher is better; unit: PERCENT. - [Multi-Semantic AQ](): Higher is better; unit: PERCENT. ## Recorded results | Method | Overall Score | Result date | Report paper | | --- | --- | --- | --- | | Nano Banana Pro | 48.35 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | SeeDream 4.5 | 44.17 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | Nano Banana | 42.88 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | GPT-Image | 40.94 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | Emu3.5-Image | 28.8 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | FLUX.2-dev | 27.37 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | Qwen-Image | 25.67 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | Ours | 23.91 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | Omini-Gen2 | 21.12 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | Bagel | 18.97 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | SeeDream 4.0 | 17.74 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | GLM-Image | 17.45 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | | NextStep-1 | 9.9 | Not recorded | 698d5ea8c97a0b3e3b5e3e7e | ### Nano Banana Pro Overall Score: 48.35. Result ID: 6998c18c5698665ec424b499. [Reporting paper]() Recorded evaluation configuration: {"Interleaved":true}. This record does not assert independent reproduction. ### SeeDream 4.5 Overall Score: 44.17. Result ID: 6998c18c5698665ec424b49d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. ### Nano Banana Overall Score: 42.88. Result ID: 6998c18c5698665ec424b49a. [Reporting paper]() Recorded evaluation configuration: {"Interleaved":true}. This record does not assert independent reproduction. ### GPT-Image Overall Score: 40.94. Result ID: 6998c18c5698665ec424b49b. [Reporting paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. ### Emu3.5-Image Overall Score: 28.8. Result ID: 6998c18c5698665ec424b4a2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. ### FLUX.2-dev Overall Score: 27.37. Result ID: 6998c18c5698665ec424b4a0. [Reporting paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. ### Qwen-Image Overall Score: 25.67. Result ID: 6998c18c5698665ec424b49e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. ### Ours Overall Score: 23.91. Result ID: 6998c18c5698665ec424b4a5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Interleaved":true}. This record does not assert independent reproduction. ### Omini-Gen2 Overall Score: 21.12. Result ID: 6998c18c5698665ec424b4a3. [Reporting paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. ### Bagel Overall Score: 18.97. Result ID: 6998c18c5698665ec424b4a4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Interleaved":true}. This record does not assert independent reproduction. ### SeeDream 4.0 Overall Score: 17.74. Result ID: 6998c18c5698665ec424b49c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. ### GLM-Image Overall Score: 17.45. Result ID: 6998c18c5698665ec424b49f. [Reporting paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. ### NextStep-1 Overall Score: 9.9. Result ID: 6998c18c5698665ec424b4a1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Interleaved":false}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.