Concept art & fantasy
How 22 AI image models render: “An alien restaurant on a strange planet. The interior is well lit, but the décor is very strange and out of this world. Alien waitresses are bringing in platters of food. Alien diners are eating very unusual dishes.”
Same prompt, sent verbatim, at each model's default settings and a fixed seed. Judge: design originality, readable silhouette, lighting drama. 72 community votes so far.
Community prompt by @philbland · voting closed 2026-10-08
22 text-to-image models rendered this concept art & fantasy prompt with the identical text at their default settings and a fixed seed. GPT Image 2 is the community favourite so far, winning 67% of its 6 head-to-head votes on this prompt. Grok Imagine Image Quality was the fastest at 6.3s.
What to look for: sense of scale, atmosphere and light, and designs that read clearly at a glance.
- GPT Image 267%OpenAI63.0s · 42 cr Try
- Seedream 5.0 Flash63%ByteDance28.8s · 22 cr Try
- Seedream 5.0 Pro60%ByteDance73.8s · 49 cr Try
- GPT Image 2.5 Sunburst57%OpenAI43.0s · 40 cr Try
- GPT Image 2.5 Flare56%OpenAI29.7s · 40 cr Try
- Ideogram v450%Ideogram19.4s · 40 cr Try
- Krea 2 Large50%Krea103.2s · 80 cr Try
- 50%22.1s · 50 cr Try
- Nano Banana 2 Lite50%Google12.8s · 36 cr Try
- Ideogram v4.550%Ideogram15.2s · 80 cr Try
- ImagineArt 245%ImagineArt51.6s · 60 cr Try
- Qwen Image 3.0 Pro45%Alibaba10.4s · 36 cr Try
- 6.3s · 80 cr Try
- Qwen Image 3.029%Alibaba7.5s · 32 cr Try
- Krea 2 Medium25%Krea15.2s · 40 cr Try
- Kling v3 Omni22%Kuaishou70.1s · 35 cr Try
- FLUX 3 ImageBlack Forest Labs21.9s · 65 cr Try
- ImagineArt 1.5 ProImagineArt · retired from voting52.4s · 55 cr Try
- Recraft v4 ProRecraft · retired from voting14.6s · 300 cr Try
- Muse Image 1.0Meta14.5s · 12 cr Try
- Nano Banana 2.1Google10.7s · 40 cr Try
FAQ
- Which AI model renders this prompt best?
- GPT Image 2 leads this prompt in the BudgetPixel AI Image Arena, winning 67% of its 6 blind head-to-head votes. Vote on the arena page to add your judgement.
- Were the prompts adjusted per model?
- No. Every model received the exact same prompt, unmodified, at its default settings, with no negative prompt or style suffix. Some models rewrite prompts internally, and that is part of what is being compared.
- Can I run this prompt myself?
- Yes. Each tile’s “Try” link opens the BudgetPixel workshop with this prompt and that model preselected.