---
title: "Text to image leaderboard"
description: "Rankings of text-to-image models."
canonical: https://pirouter.ai/rankings/t2i
last-updated: 2026-10-15
---

# Text to image leaderboard

> Rankings of text-to-image models.

- Updated: 2026-09-24
- 9,793,824 votes · 78 models
- Score · higher is better

## Full ranking

| Rank | Rank spread | Model | Vendor | Score | Votes |
|---|---|---|---|---|---|
| 1 | 1–1 | GPT-image 2.5 Sunburst | OpenAI | 1424 ±8 | 10884 |
| 2 | 2–2 | GPT-image 2.5 Flare | OpenAI | 1401 ±8 | 10058 |
| 3 | 3–3 | GPT-image 2 (medium) | OpenAI | 1383 ±4 | 88744 |
| 4 | 4–4 | MAI Image 2.6 | Microsoft AI | 1335 ±6 | 18625 |
| 5 | 5–6 | Reve 2.1 | Reve | 1302 ±8 | 8305 |
| 6 | 5–6 | Grok Imagine Image 2.0 (low) | xAI | 1301 ±8 | 7257 |
| 7 | 7–8 | Muse Image | Meta | 1276 ±5 | 35422 |
| 8 | 7–9 | Reve 2.0 | Reve | 1269 ±6 | 15802 |
| 9 | 8–12 | [Nano Banana 2](https://pirouter.ai/models/google/gemini-3.1-flash-image) | Google | 1261 ±4 | 53228 |
| 10 | 9–13 | Seedream 5.0 Pro | ByteDance | 1256 ±4 | 85690 |
| 11 | 9–14 | Qwen Image 3.0 Pro | Alibaba | 1256 ±6 | 11520 |
| 12 | 9–14 | MAI Image 2.5 | Microsoft AI | 1254 ±4 | 61448 |
| 13 | 10–14 | [Nano Banana 2 Lite](https://pirouter.ai/models/google/gemini-3.1-flash-lite-image) | Google | 1250 ±6 | 17851 |
| 14 | 11–14 | [Nano Banana Pro](https://pirouter.ai/models/google/gemini-3-pro-image) | Google | 1247 ±3 | 168447 |
| 15 | 15–16 | GPT-image 1.5 High Fidelity | OpenAI | 1239 ±3 | 165437 |
| 16 | 15–16 | Qwen Image 2.1 | Alibaba | 1228 ±9 | 4445 |
| 17 | 17–17 | Ideogram 4.0 Quality | Ideogram | 1205 ±4 | 53371 |
| 18 | 18–20 | Qwen Image 2.0 Pro 2026 06 22 | Alibaba | 1192 ±6 | 13123 |
| 19 | 18–20 | Uni 1.1 (max) | Luma | 1189 ±6 | 14083 |
| 20 | 18–21 | MAI Image 2 | Microsoft AI | 1184 ±5 | 51477 |
| 21 | 21–23 | Grok Imagine Image | xAI | 1170 ±3 | 262132 |
| 22 | 20–27 | Recraft V4.1 Utility Pro | Recraft | 1169 ±11 | 2618 |
| 23 | 21–27 | Cosmos3 Super Text2image | NVIDIA | 1165 ±7 | 10856 |
| 24 | 22–27 | Grok Imagine Image Pro | xAI | 1162 ±4 | 99444 |
| 25 | 22–27 | Flux 2 (max) | Black Forest Labs | 1162 ±4 | 123187 |
| 26 | 22–29 | Flux 2 Flex | Black Forest Labs | 1157 ±3 | 156912 |
| 27 | 22–32 | Reve V1.5 | Reve | 1155 ±4 | 39755 |
| 28 | 26–32 | Flux 2 Pro | Black Forest Labs | 1154 ±3 | 205540 |
| 29 | 26–35 | Cosmos3 Super Text2image | NVIDIA | 1152 ±6 | 18581 |
| 30 | 27–34 | Gemini 2.5 Flash Image Preview | Google | 1150 ±2 | 878451 |
| 31 | 27–35 | Hunyuan Image 3.0 | Tencent | 1150 ±3 | 180550 |
| 32 | 27–35 | Imagen Ultra 4.0 Generate 001 | Google | 1148 ±4 | 388772 |
| 33 | 29–35 | Seedream 4.5 | ByteDance | 1148 ±3 | 287128 |
| 34 | 29–36 | Flux 2 Dev | Black Forest Labs | 1146 ±4 | 78080 |
| 35 | 30–38 | Seedream 4.2k | ByteDance | 1140 ±7 | 12566 |
| 36 | 35–38 | Seedream 5.0 Lite | ByteDance | 1138 ±3 | 123178 |
| 37 | 35–38 | Wan2.6 T2i | Alibaba | 1137 ±3 | 223535 |
| 38 | 34–42 | Recraft V4.1 Pro | Recraft | 1133 ±10 | 2795 |
| 39 | 38–41 | Imagen 4.0 Generate 001 | Google | 1129 ±3 | 539788 |
| 40 | 38–42 | Qwen Image 2512 | Alibaba | 1125 ±3 | 98363 |
| 41 | 38–43 | Krea 2 (medium) | Krea | 1123 ±4 | 50800 |
| 42 | 39–50 | Seedream 4 Fal | ByteDance | 1117 ±7 | 11842 |
| 43 | 42–49 | Wan2.5 T2i Preview | Alibaba | 1116 ±3 | 271922 |
| 44 | 41–50 | Hidream O1 Image | HiDream | 1116 ±4 | 59511 |
| 45 | 42–50 | GPT-image 1 | OpenAI | 1116 ±3 | 268466 |
| 46 | 42–50 | Recraft V4 | Recraft | 1115 ±3 | 129674 |
| 47 | 42–50 | Seedream 4 High Res Fal | ByteDance | 1113 ±3 | 182669 |
| 48 | 42–52 | Krea 2 Turbo | Krea | 1113 ±5 | 43913 |
| 49 | 42–53 | Krea 2 Large | Krea | 1110 ±4 | 49906 |
| 50 | 43–52 | GPT-image 1 Mini | OpenAI | 1110 ±3 | 171058 |
| 51 | 48–53 | Wan2.7 Image Pro | Alibaba | 1104 ±5 | 30229 |
| 52 | 50–54 | Wan2.7 Image | Alibaba | 1101 ±5 | 30621 |
| 53 | 48–54 | Recraft V4.1 Flash | Recraft | 1100 ±9 | 3533 |
| 54 | 52–55 | MAI Image 1 | Microsoft AI | 1093 ±4 | 98414 |
| 55 | 54–56 | Z Image Turbo | Alibaba | 1084 ±5 | 24291 |
| 56 | 55–57 | Seedream 3 | ByteDance | 1082 ±5 | 36890 |
| 57 | 56–58 | Flux 1 Kontext (max) | Black Forest Labs | 1074 ±3 | 65517 |
| 58 | 57–58 | Flux 2 Klein 9b | Black Forest Labs | 1070 ±3 | 150780 |
| 59 | 59–62 | Qwen Image Prompt Extend | Alibaba | 1060 ±3 | 716829 |
| 60 | 59–62 | Flux 1 Kontext Pro | Black Forest Labs | 1059 ±3 | 332376 |
| 61 | 59–62 | Imagen 3.0 Generate 002 | Google | 1058 ±3 | 359786 |
| 62 | 59–62 | Qwen Image | Alibaba | 1057 ±3 | 84555 |
| 63 | 63–63 | Ideogram V3 Quality | Ideogram | 1048 ±4 | 118542 |
| 64 | 64–66 | Photon | Luma | 1036 ±4 | 130158 |
| 65 | 64–66 | P Image | — | 1033 ±3 | 109782 |
| 66 | 64–67 | Flux 2 Klein 4b | Black Forest Labs | 1030 ±3 | 152358 |
| 67 | 66–69 | Runway Gen4 | Runway | 1024 ±4 | 54376 |
| 68 | 67–70 | Recraft V3 | Recraft | 1021 ±4 | 195623 |
| 69 | 68–72 | Flux 1.1 Pro | Black Forest Labs | 1016 ±4 | 70460 |
| 70 | 69–72 | Ideogram V2 | Ideogram | 1013 ±4 | 72090 |
| 71 | 69–72 | Lucid Origin | Leonardo AI | 1013 ±3 | 287345 |
| 72 | 67–72 | GLM Image | Z.ai | 1012 ±9 | 4852 |
| 73 | 73–74 | Gemini 2.0 Flash Preview Image Generation | Google | 975 ±3 | 257241 |
| 74 | 73–75 | Flux 1 Dev Fp8 | Black Forest Labs | 969 ±4 | 49221 |
| 75 | 74–75 | Dall E 3 | OpenAI | 968 ±4 | 238749 |
| 76 | 76–77 | Flux 1 Kontext Dev | Black Forest Labs | 941 ±4 | 216238 |
| 77 | 76–77 | Stable Diffusion V35 Large | — | 938 ±5 | 23396 |
| 78 | 78–78 | Bagel | ByteDance | 898 ±6 | 12363 |

## FAQ

### What is the best AI model for text to image?

As of 2026-09-24, OpenAI GPT-image 2.5 Sunburst is #1 on the text to image board with a score of 1424, rank spread 1–1.

### What is the best open-weights model for text to image?

Qwen Image 2.1 (Alibaba), #16 on this board.

### What does this board rank?

Rankings of text-to-image models.

### How often is it updated?

Whenever new models ship or the rankings change; this board was last updated on 2026-09-24.

Data: [LMArena leaderboard dataset](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset), used under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/); filtered and reasoning tiers merged, scores unchanged.
View the original board: https://arena.ai/leaderboard/text-to-image