Concept art & fantasy
How 21 AI image models render: “A mystical dragon's head, slender and dreamy, rests upon the yellowed pages of an ancient tome, the cover worn and mysterious, as soft warm light pours in through a nearby window, casting a serene ambiance, the air thick with the scent of old parchment and forgotten knowledge.”
Same prompt, sent verbatim, at each model's default settings and a fixed seed. Judge: design originality, readable silhouette, lighting drama. 60 community votes so far.
Community prompt by @chris70 · voting closed 2026-10-19
21 text-to-image models rendered this concept art & fantasy prompt with the identical text at their default settings and a fixed seed. GPT Image 2.5 Sunburst is the community favourite so far, winning 100% of its 5 head-to-head votes on this prompt. FLUX 3 Image Klein was the fastest at 4.5s.
What to look for: sense of scale, atmosphere and light, and designs that read clearly at a glance.
- GPT Image 2.5 Sunburst100%OpenAI44.3s · 40 cr Try
- GPT Image 2.5 Flare83%OpenAI38.1s · 40 cr Try
- Ideogram v4.580%Ideogram11.5s · 80 cr Try
- Seedream 5.0 Flash60%ByteDance18.0s · 22 cr Try
- Muse Image 1.050%Meta13.6s · 12 cr Try
- Nano Banana 2.143%Google10.1s · 31 cr Try
- Qwen Image 3.0 Pro43%Alibaba10.3s · 36 cr Try
- 8.0s · 80 cr Try
- FLUX 3 Image33%Black Forest Labs21.1s · 65 cr Try
- 20%15.0s · 50 cr Try
- Seedream 5.0 Pro20%ByteDance64.0s · 49 cr Try
- Ideogram v414%Ideogram14.3s · 40 cr Try
- Krea 2 Large14%Krea29.2s · 80 cr Try
- ImagineArt 2ImagineArt · retired from voting27.6s · 60 cr Try
- Kling v3 OmniKuaishou36.8s · 35 cr Try
- Krea 2 MediumKrea34.2s · 40 cr Try
- Nano Banana 2 LiteGoogle496.3s · 36 cr Try
- Qwen Image 3.0Alibaba7.7s · 32 cr Try
- FLUX 3 Image KleinBlack Forest Labs4.5s · 12 cr Try
- Midjourney V8.2Midjourney97.0s · 160 cr Try
FAQ
- Which AI model renders this prompt best?
- GPT Image 2.5 Sunburst leads this prompt in the BudgetPixel AI Image Arena, winning 100% of its 5 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.