Image · Text to Image · July 28, 2026
Text to Image Leaderboard
Every major text-to-image model ranked by Elo from blind head-to-head human votes — no cherry-picked grids, no lab-supplied samples.
Models ranked
45
Open weights
10
Arena appearances
419,599
Prompt categories
5
Highlights
Higher is better
Arena Elo
Overall leaderboard · best model per lab
- 1GPT Image 2 (high)1339
- 2Reve 2.11301
- 3MAI-Image-2.51271
- 4Nano Banana 2 (Gemini 3.1 Flash Image Preview)1263
- 5Riverflow 2.01254
- 6HiDream-O1-Image-1.51247
- 7Seedream 5.0 Pro1239
- 8Cosmos3-Super-Text2Image (agentic)1218
- 9grok-imagine-image-quality1204
Higher is better
Open Weights
Downloadable checkpoints · best per lab
- 1Cosmos3-Super-Text2Image (agentic)1218
- 2HiDream-O1-Image-Dev-26041189
- 3Ideogram 4.0 (Quality)1175
- 4ERNIE Image1168
- 5Ideogram 4.0 Fast (Quality)1158
- 6Qwen Image Max 25121157
- 7FLUX.2 [dev]1154
Lower is better
Price per Image
USD per generation, default quality · cheapest per lab
- 1FLUX.2 [dev] Turbo$0.01
- 2ERNIE Image Turbo$0.01
- 3FLUX.2 [dev]$0.01
- 4Krea 2 Medium Turbo$0.01
- 5MAI-Image-2.5-Flash$0.02
- 6grok-imagine-image$0.02
- 7Qwen Image Max 2512$0.02
- 8Reve 2.1$0.02
- 9Seedream 4.0$0.03
Global leaderboard
Elo from blind head-to-head votes · All categories
| Rank | Model | Creator | Elo | 95% CI | Appearances | $/Image | Released |
|---|---|---|---|---|---|---|---|
| 1 | GPT Image 2 (high) | OpenAI | 1339 | ±7 | 21,290 | $0.21 | Apr 2026 |
| 2 | Reve 2.1 | Reve | 1301 | ±11 | 9,634 | $0.02 | Jul 2026 |
| 3 | MAI-Image-2.5 | Microsoft AI | 1271 | ±9 | 9,683 | $0.05 | Jun 2026 |
| 4 | Nano Banana 2 (Gemini 3.1 Flash Image Preview) | 1263 | ±7 | 19,728 | $0.07 | Feb 2026 | |
| 5 | Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) | 1262 | ±9 | 9,648 | $0.03 | Jun 2026 | |
| 6 | GPT Image 1.5 (high) | OpenAI | 1262 | ±7 | 12,563 | $0.13 | Dec 2025 |
| 7 | Riverflow 2.0 | Sourceful | 1254 | ±8 | 7,850 | $0.15 | Feb 2026 |
| 8 | HiDream-O1-Image-1.5 | HiDream | 1247 | ±8 | 12,490 | $0.08 | Jun 2026 |
| 9 | Reve 2.0 | Reve | 1243 | ±10 | 3,740 | $0.02 | Jun 2026 |
| 10 | Seedream 5.0 Pro | Bytedance | 1239 | ±9 | 8,263 | $0.09 | Jul 2026 |
| 11 | Nano Banana Pro (Gemini 3 Pro Image) | 1225 | ±7 | 11,893 | $0.13 | Nov 2025 | |
| 12 | Cosmos3-Super-Text2Image (agentic)Open | NVIDIA | 1218 | ±8 | 10,044 | — | May 2026 |
| 13 | MAI-Image-2.5-Flash | Microsoft AI | 1208 | ±9 | 4,992 | $0.02 | Jun 2026 |
| 14 | grok-imagine-image-quality | xAI | 1204 | ±7 | 15,888 | $0.05 | Apr 2026 |
| 15 | Recraft V4.1 Utility | Recraft | 1204 | ±8 | 5,543 | $0.04 | May 2026 |
| 16 | Recraft V4.1 Utility Pro | Recraft | 1201 | ±7 | 15,319 | $0.21 | May 2026 |
| 17 | FLUX.2 [max] | Black Forest Labs | 1196 | ±7 | 11,649 | $0.07 | Dec 2025 |
| 18 | Seedream 4.0 | ByteDance Seed | 1192 | ±7 | 17,190 | $0.03 | Sep 2025 |
| 19 | Krea 2 Medium | Krea | 1190 | ±7 | 15,174 | $0.03 | May 2026 |
| 20 | Krea 2 Large | Krea | 1190 | ±7 | 15,536 | $0.06 | May 2026 |
| 21 | HiDream-O1-Image-Dev-2604Open | HiDream | 1189 | ±7 | 7,226 | — | May 2026 |
| 22 | MAI-Image-2 | Microsoft AI | 1188 | ±7 | 7,816 | $0.04 | Mar 2026 |
| 23 | Cosmos3-Super-Text2Image-4StepOpen | NVIDIA | 1188 | ±10 | 5,275 | — | Jul 2026 |
| 24 | FLUX.2 [pro] | Black Forest Labs | 1187 | ±7 | 10,970 | $0.03 | Nov 2025 |
| 25 | Krea 2 Medium Turbo | Krea | 1187 | ±10 | 3,076 | $0.01 | Jun 2026 |
| 26 | grok-imagine-image | xAI | 1181 | ±7 | 9,772 | $0.02 | Jan 2026 |
| 27 | Imagen 4 Ultra Preview 0606 | 1181 | ±8 | 5,953 | $0.06 | Jun 2025 | |
| 28 | FLUX.2 [flex] | Black Forest Labs | 1180 | ±8 | 6,462 | $0.06 | Nov 2025 |
| 29 | Luma UNI 1 Max | Luma Labs | 1175 | ±8 | 5,786 | $0.10 | May 2026 |
| 30 | Ideogram 4.0 (Quality)Open | Ideogram | 1175 | ±8 | 10,085 | $0.10 | Jun 2026 |
| 31 | Qwen Image 2.0 Pro (2026-04-22) | Alibaba | 1174 | ±8 | 5,848 | $0.07 | Apr 2026 |
| 32 | Imagen 4 Ultra | 1172 | ±8 | 6,822 | $0.06 | Jun 2025 | |
| 33 | ImagineArt 2.0 | ImagineArt | 1169 | ±7 | 10,990 | $0.03 | Apr 2026 |
| 34 | Seedream 4.5 | ByteDance Seed | 1169 | ±7 | 12,981 | $0.04 | Dec 2025 |
| 35 | ERNIE ImageOpen | Baidu | 1168 | ±8 | 4,871 | $0.03 | Apr 2026 |
| 36 | ERNIE Image TurboOpen | Baidu | 1163 | ±8 | 5,764 | $0.01 | Apr 2026 |
| 37 | Imagen 4 Preview 0606 | 1162 | ±8 | 5,868 | $0.04 | Jun 2025 | |
| 38 | MAI-Image-2-Efficient | Microsoft AI | 1158 | ±7 | 8,412 | $0.02 | Apr 2026 |
| 39 | Ideogram 4.0 Fast (Quality)Open | Fal | 1158 | ±10 | 3,661 | $0.02 | Jul 2026 |
| 40 | Recraft V4.1 Pro | Recraft | 1158 | ±8 | 5,122 | $0.21 | May 2026 |
| 41 | Qwen Image Max 2512Open | Alibaba | 1157 | ±8 | 5,836 | $0.02 | Dec 2025 |
| 42 | FLUX.2 [dev] TurboOpen | Fal | 1154 | ±8 | 6,399 | $0.01 | Dec 2025 |
| 43 | Nano Banana (Gemini 2.5 Flash Image) | 1154 | ±7 | 11,175 | $0.04 | Aug 2025 | |
| 44 | FLUX.2 [dev]Open | Black Forest Labs | 1154 | ±7 | 9,866 | $0.01 | Nov 2025 |
| 45 | Recraft V4.1 | Recraft | 1153 | ±8 | 5,446 | $0.04 | May 2026 |
Showing 45 of 45 models. Category Elo is computed from votes on that prompt type alone.
Arena Elo — Open Weights
Every downloadable checkpoint on the board, ranked by overall Elo. Higher is better.
Cosmos3 sits fifth overall — ahead of most closed APIs — and the FLUX.2 [dev] line keeps Black Forest Labs in four of the fourteen open slots.
Arena Elo vs. Price per Image
Crowd-judged quality against USD per generation (log scale).
Up and to the left wins. The open-weights mid-table — FLUX.2 [dev], Qwen-Image-2512, Hunyuan — is where the value lives; models without public per-image pricing are excluded.
Frontier Elo Over Time
Arena Elo by release date. The dashed line tracks the running frontier.
The frontier has changed hands five times since Imagen 4 Ultra Preview 0606 — Google, ByteDance Seed, and OpenAI have all held it. GPT Image 2 set the current mark in April 2026.
How the arena works
1 · Prompt
Prompts are sampled across five categories — people, photorealism, art, product, and typography — so no model can farm a single style.
2 · Blind matchup
Two models generate from the same prompt. Voters see the images side by side with no names, no logos, and no way to tell who made what.
3 · Elo
Each vote updates both models' Elo ratings. Confidence intervals come from resampling the vote history; new models stay provisional until they clear roughly 3,000 appearances.
FAQ
Which model leads the arena right now?
GPT Image 2 (high) holds the top spot with an Elo of 1339 across 21,290 appearances — a 38-point gap over the chasing pack.
And among open weights?
NVIDIA's Cosmos3-Super-Text2Image (agentic) leads the open-weights cut at 1218 — twelfth overall, ahead of plenty of closed APIs — followed by HiDream-O1-Image-Dev-2604 (1189) and NVIDIA's Cosmos3-Super-Text2Image-4Step (1188).
How are models ranked?
Pure crowd judgment. Voters get a prompt and two unlabeled images from different models and pick the better one; thousands of those blind matchups feed an Elo rating, with a 95% confidence interval from resampling the vote history.
What counts as open weights?
The checkpoint has to be downloadable and runnable on your own hardware under a published license. API-only siblings of open families — FLUX.2 [pro], Qwen Image 2.0 Pro — stay in the proprietary column.
Latest Research
Reporting from the paper mines
Mark Zuckerberg blasts centralisation of AI Power
By The Straits Times · 45 mins ago
SK Telecom Unveils 688-Billion-Parameter AI Model A.X K2
By 조선일보 · 50 mins ago
AI boom will sustain Hong Kong's export surge, lift GDP outlook, economists say
By South China Morning Post · 1 hr ago
Best AI Financial Research Platforms for Smarter Investing in 2026
By Analytics Insight · 1 hr ago
Methodology: rankings come from blind pairwise human votes; Elo, confidence intervals, and appearance counts reflect the full vote history at snapshot time. Arena data: public image-arena snapshot, July 28, 2026. Prices are list API prices per image at default quality and change frequently.