---
title: "Text to image: model rankings and source evidence · Undominated.ai"
canonical: https://undominated.ai/rankings/arena-text-to-image/
description: "Text to image ranked by Arena score. Inspect exact model variants, source dates, measurement limits and downloadable evidence."
---

# Text to image: model rankings and source evidence · Undominated.ai

> Text to image ranked by Arena score. Inspect exact model variants, source dates, measurement limits and downloadable evidence.

[All rankings](/rankings/) / Images

# Text to image

Human preference ratings for the overall category of this Arena task. Scores from different boards are not interchangeable.

**84 entries** · Arena score · Higher is better

Source publication: 2026-11-04. Retrieved 2026-11-05.

Source: [Arena (lmarena-ai)](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset) · [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/).

Ranks use confidence bounds. Several models can share a rank; the rows within a shared rank are displayed by score.

Text to image · Arena score

| Significance rank | Model / evaluated variant | Arena score | 95% confidence interval | Votes |
| --- | --- | --- | --- | --- |
| 1 | GPT-Image-2.5 Sunburst openai | 1,427.13 | 1,421.89 – 1,432.37 | 54,913 |
| 2 | GPT-Image-2.6 Sunburst (Medium) openai | 1,409 | 1,396.19 – 1,421.8 | 2,967 |
| 2 | GPT-Image-2.5 Flare openai | 1,399.44 | 1,394.28 – 1,404.61 | 53,925 |
| 3 | GPT-Image-2.6 Flare (Medium) openai | 1,383.57 | 1,370.99 – 1,396.15 | 2,792 |
| 4 | GPT-Image-2 (Medium) openai | 1,382.1 | 1,378.25 – 1,385.94 | 106,737 |
| 6 | MAI-Image-2.6 microsoft-ai | 1,330.58 | 1,325.27 – 1,335.89 | 38,071 |
| 6 | Gemini Nano Banana 2.1 google | 1,325.31 | 1,316.48 – 1,334.14 | 5,239 |
| 7 | Grok Imagine Image 2.0 xai | 1,317.31 | 1,310.17 – 1,324.46 | 10,830 |
| 9 | Reve 2.1 reve | 1,300.4 | 1,292.9 – 1,307.91 | 8,237 |
| 9 | Grok Imagine Image 2.0 (20260801) xai | 1,290.74 | 1,285.35 – 1,296.14 | 30,734 |
| 11 | Muse Image meta | 1,268.12 | 1,263.56 – 1,272.68 | 57,568 |
| 11 | Reve 2.0 reve | 1,267.74 | 1,261.55 – 1,273.93 | 15,756 |
| 13 | Gemini 3.1 Flash Image (nano-banana-2, Web Search) google | 1,257.49 | 1,253.63 – 1,261.36 | 77,495 |
| 13 | Qwen-Image-3.0-Pro alibaba | 1,254.06 | 1,247.59 – 1,260.54 | 11,412 |
| 13 | MAI-Image-2.5 microsoft-ai | 1,253 | 1,248.82 – 1,257.17 | 61,222 |
| 13 | Seedream 5.0 Pro bytedance | 1,250.8 | 1,246.95 – 1,254.65 | 109,680 |
| 13 | Gemini 3.1 Flash-Lite Image (nano-banana-2-lite) google | 1,249.44 | 1,243.47 – 1,255.4 | 17,787 |
| 14 | Gemini 3 Pro Image (nano-banana-pro, 2K) google | 1,246.31 | 1,243.36 – 1,249.27 | 191,257 |
| 19 | GPT-Image-1.5 (High Fidelity) openai | 1,236.39 | 1,233.38 – 1,239.41 | 173,798 |
| 19 | Gemini 3 Pro Image Preview (nano-banana-pro) google | 1,231.79 | 1,226.91 – 1,236.67 | 86,446 |
| 20 | Qwen-Image-2.1 alibaba | 1,226.55 | 1,220.52 – 1,232.59 | 19,378 |
| 22 | Ideogram 4.0 (Quality) ideogram | 1,199.24 | 1,195.47 – 1,203 | 77,424 |
| 22 | Qwen-Image-2.0-Pro (2026-06-22) alibaba | 1,191.63 | 1,185.78 – 1,197.48 | 13,081 |
| 23 | UNI-1.1 Max luma-ai | 1,188.75 | 1,182.59 – 1,194.91 | 14,079 |
| 23 | MAI-Image-2 microsoft-ai | 1,183.36 | 1,178.81 – 1,187.91 | 51,435 |
| 23 | UNI-1.1 luma-ai | 1,182.68 | 1,178.54 – 1,186.82 | 58,567 |
| 27 | Grok Imagine Image xai | 1,169.42 | 1,166.79 – 1,172.05 | 292,345 |
| 25 | Recraft V4.1 Utility Pro recraft | 1,168.73 | 1,157.77 – 1,179.69 | 2,622 |
| 27 | Cosmos3-Super-Text2Image (Agentic) nvidia | 1,164.75 | 1,157.69 – 1,171.81 | 10,750 |
| 28 | FLUX.2 [max] bfl | 1,161.46 | 1,157.95 – 1,164.98 | 123,131 |
| 28 | Grok Imagine Image Pro xai | 1,161.29 | 1,157.53 – 1,165.05 | 99,370 |
| 28 | FLUX.2 [flex] bfl | 1,156.26 | 1,152.89 – 1,159.64 | 156,849 |
| 28 | Reve v1.5 reve | 1,154.43 | 1,150.21 – 1,158.65 | 39,684 |
| 32 | FLUX.2 [pro] bfl | 1,152.68 | 1,149.89 – 1,155.48 | 226,242 |
| 32 | Cosmos3-Super-Text2Image nvidia | 1,151.1 | 1,145.25 – 1,156.95 | 18,370 |
| 32 | HunyuanImage 3.0 tencent | 1,150.91 | 1,147.59 – 1,154.22 | 180,461 |
| 33 | Gemini 2.5 Flash Image Preview (nano-banana) google | 1,148.64 | 1,146.28 – 1,151 | 911,746 |
| 33 | Seedream 4.5 bytedance | 1,148.35 | 1,145.82 – 1,150.87 | 323,656 |
| 33 | Imagen 4 Ultra (generate-001) google | 1,146.15 | 1,141.84 – 1,150.46 | 388,699 |
| 35 | FLUX.2 [dev] bfl | 1,145.24 | 1,141.09 – 1,149.39 | 77,924 |
| 39 | Seedream 5.0 Lite bytedance | 1,140.75 | 1,137.63 – 1,143.87 | 145,323 |
| 39 | Seedream 4.0 (2K) bytedance | 1,138.56 | 1,132.01 – 1,145.11 | 12,563 |
| 41 | Wan 2.6 alibaba | 1,137.97 | 1,135.21 – 1,140.73 | 253,190 |
| 39 | Recraft V4.1 Pro recraft | 1,131.93 | 1,121.45 – 1,142.4 | 2,798 |
| 43 | Krea 2 Medium krea | 1,128.92 | 1,124.97 – 1,132.87 | 74,418 |
| 44 | Imagen 4 (generate-001) google | 1,128.35 | 1,125.63 – 1,131.08 | 539,617 |
| 44 | Qwen-Image (2512) alibaba | 1,124.67 | 1,121.22 – 1,128.12 | 98,252 |
| 48 | Wan 2.5 Preview alibaba | 1,116.62 | 1,113.88 – 1,119.36 | 293,738 |
| 46 | Seedream 4.0 (fal) bytedance | 1,116.19 | 1,109.66 – 1,122.72 | 11,841 |
| 48 | Krea 2 Large krea | 1,115.37 | 1,111.38 – 1,119.37 | 71,664 |
| 48 | GPT-Image-1 openai | 1,115.13 | 1,111.97 – 1,118.29 | 268,408 |
| 48 | Recraft V4 recraft | 1,114.8 | 1,111.7 – 1,117.9 | 154,031 |
| 48 | HiDream-O1-Image hidream | 1,114.71 | 1,111.09 – 1,118.33 | 79,970 |
| 48 | Krea 2 Turbo krea | 1,114.56 | 1,110.34 – 1,118.78 | 67,389 |
| 48 | Seedream 4.0 (High Res, fal) bytedance | 1,112.72 | 1,109.49 – 1,115.96 | 182,605 |
| 49 | GPT-Image-1 mini openai | 1,109.33 | 1,106.14 – 1,112.51 | 170,972 |
| 56 | Wan 2.7 Image Pro alibaba | 1,103.03 | 1,098 – 1,108.06 | 30,216 |
| 57 | Wan 2.7 Image alibaba | 1,100.54 | 1,095.53 – 1,105.55 | 30,626 |
| 56 | Recraft V4.1 Flash recraft | 1,098.49 | 1,088.89 – 1,108.08 | 3,438 |
| 58 | MAI-Image-1 microsoft-ai | 1,092.44 | 1,088.81 – 1,096.06 | 98,357 |
| 61 | Z-Image-Turbo alibaba | 1,082.19 | 1,076.87 – 1,087.51 | 24,923 |
| 61 | Seedream 3.0 bytedance | 1,081.53 | 1,076.46 – 1,086.61 | 36,894 |
| 61 | FLUX.1 Kontext [max] bfl | 1,075.3 | 1,072.06 – 1,078.54 | 65,502 |
| 63 | FLUX.2 [klein] 9B bfl | 1,069.84 | 1,066.63 – 1,073.05 | 150,686 |
| 65 | Qwen-Image (Prompt Extend) alibaba | 1,060.81 | 1,058.12 – 1,063.5 | 716,605 |
| 65 | FLUX.1 Kontext [pro] bfl | 1,058.85 | 1,055.42 – 1,062.29 | 332,345 |
| 65 | Imagen 3 (generate-002) google | 1,057.16 | 1,053.92 – 1,060.39 | 359,756 |
| 65 | Qwen-Image alibaba | 1,057 | 1,054.37 – 1,059.63 | 84,535 |
| 69 | Ideogram 3.0 (Quality) ideogram | 1,048.57 | 1,044.79 – 1,052.35 | 118,540 |
| 70 | Luma Photon luma-ai | 1,036.02 | 1,032.28 – 1,039.76 | 130,178 |
| 70 | P-Image Organization not supplied | 1,032.41 | 1,028.93 – 1,035.88 | 109,688 |
| 70 | FLUX.2 [klein] 4B bfl | 1,029.81 | 1,026.58 – 1,033.03 | 152,284 |
| 72 | Runway Gen-4 runway | 1,023.85 | 1,019.42 – 1,028.28 | 54,318 |
| 73 | Recraft V3 recraft | 1,021.64 | 1,017.96 – 1,025.33 | 195,586 |
| 74 | FLUX1.1 [pro] bfl | 1,015.69 | 1,012.21 – 1,019.18 | 70,460 |
| 75 | Ideogram 2.0 ideogram | 1,013.41 | 1,009.86 – 1,016.95 | 72,090 |
| 75 | Lucid Origin leonardo-ai | 1,013.04 | 1,009.64 – 1,016.44 | 287,317 |
| 73 | GLM-Image zai | 1,012.16 | 1,003.3 – 1,021.02 | 4,856 |
| 79 | Gemini 2.0 Flash Preview Image Generation google | 975.78 | 972.34 – 979.23 | 257,202 |
| 79 | FLUX.1 [dev] (FP8) bfl | 969.46 | 965.73 – 973.19 | 49,221 |
| 80 | DALL-E 3 openai | 967.6 | 963.88 – 971.33 | 238,732 |
| 82 | FLUX.1 Kontext [dev] bfl | 941.48 | 937.7 – 945.26 | 216,218 |
| 82 | Stable Diffusion 3.5 Large Organization not supplied | 938.35 | 933.87 – 942.83 | 23,396 |
| 84 | BAGEL bytedance | 897.32 | 891.55 – 903.1 | 12,361 |

## What this ranking can tell you

 - The population is the models in this source board, including variants and models without an available API price.
- Ranks use the published confidence bounds: one plus the number of models whose lower bound exceeds this model’s upper bound. Shared ranks are not proof of equal performance.

## Check the source

The original evidence is retained byte for byte. The downloadable board includes the source URLs, revision and SHA-256 hashes.

[Download this board and provenance →](/data/rankings/arena-text-to-image.json)

 Source files and licences - [metadata · a7a159f557a1](/data/rankings/evidence/a7a159f557a1e594a5d41218fe8d53c63c63daa39b655e51d6fc5b482a17d059.json) · [upstream source ↗](https://huggingface.co/api/datasets/lmarena-ai/leaderboard-dataset)
- [licence · 88a46d472f35](/data/rankings/evidence/88a46d472f35274fb6b7a852c3d16c8ee8007d51b37f290161537bec26895ee4.txt) · [upstream source ↗](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/raw/8b6deb69ace01ee8dcda9e46eba7a5e8a808ec53/README.md)
- [rows · cd9d3141c9a9](/data/rankings/evidence/cd9d3141c9a91a877d20ae1e4ba2b530bfad4db747242dbe706cd9fd492c4306.parquet) · [upstream source ↗](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/8b6deb69ace01ee8dcda9e46eba7a5e8a808ec53/text_to_image/latest-00000-of-00001.parquet)

[Explore other model types →](/rankings/)[Read the benchmark guide →](/benchmarks/)

## Continue your investigation

 - [Understand the measurements](/benchmarks/)
- [Compare task performance](/best/)
- [Compare embedding models](/embeddings/)
- [Inspect ranking rules](/methodology/)