# DesignArena Image generation

> Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/designarena%3Aimage. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/designarena%3Aimage.

Human pairwise preference evaluation of image generation. Voters compare model outputs for the same request; published Elo ratings and separate model configurations are retained for this specific arena and input modality.

Canonical URL: https://benchize.com/benchmarks/designarena%3Aimage
Last updated: Wed, 21 Oct 2026 18:52:20 GMT

Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations.

Human pairwise preference evaluation of image generation. Voters compare model outputs for the same request; published Elo ratings and separate model configurations are retained for this specific arena and input modality.

Multimodal · Models tested: 23 · Score publishers: [designarena](https://www.designarena.ai/leaderboard/image)

## Model results

| # | Model | Score |  |
| --- | --- | --- | --- |
| 1 | [GPT Image 2.5 Sunburst](https://benchize.com/models/gpt-image-2-5-sunburst.md) | 1,408elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1408 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gpt-image-2.5-sunburst variationName public ### Reported statistics averageGenerationTimeMs 26777 battles 10938 losses 3243 standardError Not reported winRate 70.3 wins 7695 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label GPT-Image-2.5 Sunburst Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/0` |  |  |  |
| 2 | [GPT Image 2.5 Flare](https://benchize.com/models/gpt-image-2-5-flare.md) | 1,385elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1385 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gpt-image-2.5-flare variationName public ### Reported statistics averageGenerationTimeMs 18188 battles 11385 losses 3605 standardError Not reported winRate 68.3 wins 7780 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label GPT-Image-2.5 Flare Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/1` |  |  |  |
| 3 | [GPT Image 2](https://benchize.com/models/gpt-image-2.md) | 1,365elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1365 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gpt-image-2 variationName public ### Reported statistics averageGenerationTimeMs 61712 battles 61210 losses 17337 standardError Not reported winRate 71.7 wins 43873 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label GPT Image 2 Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/2` |  |  |  |
| 4 | [Nano Banana 2.1](https://benchize.com/models/gemini-nano-banana-2-1.md) | 1,290elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1290 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion nano-banana-2.1 variationName public ### Reported statistics averageGenerationTimeMs 20224 battles 3251 losses 1218 standardError Not reported winRate 62.5 wins 2033 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Nano Banana 2.1 Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/10` |  |  |  |
| 5 | [GPT Image 1.5](https://benchize.com/models/gpt-image-1-5.md) | 1,285elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1285 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion chestnut variationName public ### Reported statistics averageGenerationTimeMs 44510 battles 378357 losses 141313 standardError Not reported winRate 62.6 wins 237044 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label GPT-Image-1.5 Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/11` |  |  |  |
| 6 | [Nano Banana 2 (Gemini 3.1 Flash Image Preview)](https://benchize.com/models/gemini-3-1-flash-image-preview.md) | 1,269elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1269 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gemini-3.1-flash-image-preview variationName public ### Reported statistics averageGenerationTimeMs 18206 battles 91596 losses 32609 standardError Not reported winRate 64.4 wins 58987 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Gemini 3.1 Flash Image Gen (Nano Banana 2) Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/14` |  |  |  |
| 7 | [Gemini 3.1 Flash Lite Image](https://benchize.com/models/gemini-3-1-flash-lite-image.md) | 1,239elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1239 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gemini-3.1-flash-lite-image variationName public ### Reported statistics averageGenerationTimeMs 6068 battles 155246 losses 72565 standardError Not reported winRate 53.3 wins 82681 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/19` |  |  |  |
| 8 | [Nano Banana Pro (Gemini 3 Pro Image Preview)](https://benchize.com/models/gemini-3-pro-image-preview.md) | 1,216elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1216 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gemini-3-pro-image-preview variationName public ### Reported statistics averageGenerationTimeMs 18662 battles 5033 losses 1994 standardError Not reported winRate 60.4 wins 3039 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Gemini 3 Pro Image Preview Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/23` |  |  |  |
| 9 | [Krea 2 Medium](https://benchize.com/models/krea-2-medium.md) | 1,211elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1211 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion krea-2-medium variationName public ### Reported statistics averageGenerationTimeMs 22812 battles 234863 losses 119182 standardError Not reported winRate 49.3 wins 115681 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Krea 2 Medium Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/25` |  |  |  |
| 10 | [Krea 2 Large](https://benchize.com/models/krea-2-large.md) | 1,207elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1207 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion krea-2-large variationName public ### Reported statistics averageGenerationTimeMs 36912 battles 23350 losses 10962 standardError Not reported winRate 53 wins 12388 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Krea 2 Large Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/27` |  |  |  |
| 11 | [FLUX.2 Flex](https://benchize.com/models/flux-2-flex.md) | 1,200elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1200 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion flux-2-flex variationName public ### Reported statistics averageGenerationTimeMs 19788 battles 67675 losses 30310 standardError Not reported winRate 55.2 wins 37365 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label FLUX.2 [flex] Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/29` |  |  |  |
| 12 | [Flux 2 Pro](https://benchize.com/models/flux-2-pro.md) | 1,195elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1195 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion flux-2-pro variationName public ### Reported statistics averageGenerationTimeMs 17064 battles 306939 losses 156275 standardError Not reported winRate 49.1 wins 150664 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label FLUX.2 [pro] Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/30` |  |  |  |
| 13 | [Recraft V4.1 Utility Pro](https://benchize.com/models/recraft-v4-1-utility-pro.md) | 1,194elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1194 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion recraftv4_1_utility_pro_raster variationName public ### Reported statistics averageGenerationTimeMs 25923 battles 13063 losses 6109 standardError Not reported winRate 53.2 wins 6954 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Recraft V4.1 Utility Pro Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/31` |  |  |  |
| 14 | [GPT Image 1](https://benchize.com/models/gpt-image-1.md) | 1,191elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1191 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gpt-image-1 variationName public ### Reported statistics averageGenerationTimeMs 43859 battles 174927 losses 80604 standardError Not reported winRate 53.9 wins 94323 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label GPT-Image-1 Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/34` |  |  |  |
| 15 | [GPT Image 1 Mini](https://benchize.com/models/gpt-image-1-mini.md) | 1,184elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1184 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gpt-image-1-mini variationName public ### Reported statistics averageGenerationTimeMs 42399 battles 228605 losses 112021 standardError Not reported winRate 51 wins 116584 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label GPT-Image-1 Mini Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/37` |  |  |  |
| 16 | [Gemini 2.5 Flash Image](https://benchize.com/models/gemini-2-5-flash-image.md) | 1,179elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1179 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion gemini-2.5-flash-image variationName public ### Reported statistics averageGenerationTimeMs 7436 battles 190860 losses 88994 standardError Not reported winRate 53.4 wins 101866 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Gemini 2.5 Flash Image (Nano Banana) Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/40` |  |  |  |
| 17 | [Recraft V4.1 Pro](https://benchize.com/models/recraft-v4-1-pro.md) | 1,163elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1163 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion recraftv4_1_pro_raster variationName public ### Reported statistics averageGenerationTimeMs 21910 battles 12638 losses 6429 standardError Not reported winRate 49.1 wins 6209 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Recraft V4.1 Pro Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/48` |  |  |  |
| 17 | [Recraft V4 Pro](https://benchize.com/models/recraft-v4-pro.md) | 1,163elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1163 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion recraftv4_pro variationName public ### Reported statistics averageGenerationTimeMs 35408 battles 44557 losses 22406 standardError Not reported winRate 49.7 wins 22151 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Recraft V4 Pro Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/49` |  |  |  |
| 19 | [Wan 2.7](https://benchize.com/models/wan-2-7.md) | 1,159elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1159 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion wan-v2.7-t2i variationName public ### Reported statistics averageGenerationTimeMs 30667 battles 102817 losses 57104 standardError Not reported winRate 44.5 wins 45713 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Wan 2.7 Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/50` |  |  |  |
| 20 | [Recraft V4.1](https://benchize.com/models/recraft-v4-1.md) | 1,146elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1146 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion recraftv4_1 variationName public ### Reported statistics averageGenerationTimeMs 15171 battles 19767 losses 10483 standardError Not reported winRate 47 wins 9284 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Recraft V4.1 Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/54` |  |  |  |
| 21 | [Recraft V3](https://benchize.com/models/recraft-v3.md) | 1,102elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1102 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion recraftv3-02-11 variationName public ### Reported statistics averageGenerationTimeMs 10337 battles 40931 losses 23384 standardError Not reported winRate 42.9 wins 17547 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Recraft V3 Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/65` |  |  |  |
| 22 | [Qwen Image](https://benchize.com/models/qwen-image.md) | 1,076elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1076 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion qwen-image variationName public ### Reported statistics averageGenerationTimeMs 34895 battles 160631 losses 96590 standardError Not reported winRate 39.9 wins 64041 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label Qwen Image Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/71` |  |  |  |
| 23 | [FLUX.2 Klein 4B](https://benchize.com/models/flux-2-klein-4b.md) | 1,062elo |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1062 \| elo \| image \| Higher is better \| ### Configuration arenaType models category image inputModality Not reported sourceModelVersion flux-2-klein-4b variationName public ### Reported statistics averageGenerationTimeMs 5406 battles 135447 losses 83158 standardError Not reported winRate 38.6 wins 52289 ### Source [Original source](https://www.designarena.ai/api/leaderboard#board=image) Source designarena Original model label FLUX.2 Klein 4B Source version public Observed 2026-10-21T16:38:04.902Z Evidence SHA-256 `63030dd7a501effde798521b3d097112a00a2f78b4a036c40bb0e4fa7a276a2a` Source locator `/data/76` |  |  |  |

---
Cite as: DesignArena Image generation — Benchize, https://benchize.com/benchmarks/designarena%3Aimage (retrieved 2026-10-21)