Documentation Index
Fetch the complete documentation index at: https://docs.imagine.art/llms.txt
Use this file to discover all available pages before exploring further.
Dreamina Image 3.1 is part of ByteDance’s Seedream model family. For the latest version with native 4K and multi-reference support, see Seedream 4.0.
Capabilities
Broad style range
Adapts to documentary, fashion, commercial photography, Baroque, Cubism, ink wash, anime characters, and fantasy scenes.
Text rendering
Place text in quotes within your prompt for improved rendering. Supports English and Chinese text in images.
Precise style control
Claimed 98% prompt adherence for style and composition instructions — reliable for specific creative briefs.
How to use
Write your prompt
Describe the subject, environment, style, and lighting. For text in images, put the exact text in quotation marks within your prompt.
Prompting tips
- For text rendering — Put the exact text you want in the image in quotes: A poster with “SPRING COLLECTION 2025” in bold serif text.
- For cinematic style — Use descriptive lighting language: “volumetric lighting”, “Rembrandt lighting”, “neon-lit rain reflections”, “golden hour backlighting”.
- For anime or illustration — Specify art style explicitly: “anime key visual”, “ink wash painting”, “Baroque oil painting style”.
Example prompts
A woman in a crimson ball gown standing in an abandoned baroque ballroom, shafts of moonlight streaming through broken windows, atmospheric dust particles, cinematic composition.
A stylized anime character with silver hair and glowing eyes in a futuristic cityscape, neon lights, detailed illustration style.
Compare models
| Model | Cinematic style | Anatomy | Multi-reference | Best for |
|---|---|---|---|---|
| Dreamina Image 3.1 | Excellent | Improved | No | Cinematic portraits, stylized art |
| Seedream 4.0 | Excellent | Strong | Up to 6 | Commercial campaigns, 4K output |
| Seedream v3 | Strong | Good | No | Fast branded imagery |
| ImagineArt 1.5 Pro | Excellent | Strong | Multi-ref | Posters, product visuals |