Dreamina V3 is useful when a workflow needs both prompt-based image generation and light reference-image editing without exposing separate endpoints. The MakeInfluencer adapter selects the text-to-image variant when no image is supplied and the edit variant when imageUrls contains one source.
Inputs and Routing
| Input | Behavior |
|---|---|
prompt | Required for both generation and editing |
aspectRatio | Common square, landscape, portrait, and ultra-wide ratios |
imageUrls omitted | Text-to-image generation |
imageUrls with one URL | Single-reference edit |
The live Dreamina V3 API documentation publishes the exact ratios, default, and credit preview.
Text-to-Image Request
curl --request POST \
https://www.makeinfluencer.ai/api/v1/models/dreamina-v3/preview-credits \
--header "Authorization: Bearer $MAKEINFLUENCER_API_KEY" \
--header "Content-Type: application/json" \
--data '{
"input": {
"prompt": "Editorial travel creator on a quiet Mediterranean terrace, linen styling, morning sun",
"aspectRatio": "4:5"
}
}'
Reference Edit Request
{
"input": {
"prompt": "Preserve the same person and pose; change the background to a minimal warm-gray studio",
"aspectRatio": "4:5",
"imageUrls": ["https://example.com/approved-creator.jpg"]
}
}
The outer generation route stays the same. Only the model input changes.
Where Dreamina V3 Fits
Campaign Concepts
Generate polished visual directions across common social and advertising ratios.
Single-Source Adaptation
Edit one approved image without switching to another model endpoint.
Platform Reframing
Create square, portrait, landscape, and ultra-wide compositions from one workflow.
Simple Product UX
Expose a single model while routing generation and editing from the presence of a source image.
Dreamina is especially convenient for applications that want one “optional reference image” control. If the user uploads nothing, the app generates. If the user uploads one image, the same model performs an edit.
Dreamina V3 vs Qwen Image 2 Pro Edit
Use Dreamina V3 for a flexible generation-plus-editing experience with one optional source. Use Qwen Image 2 Pro Edit when the job is explicitly a precision edit or needs several references.
| Requirement | Dreamina V3 | Qwen Image 2 Pro Edit |
|---|---|---|
| Text-to-image | Yes | No, editing-focused |
| Optional source image | One | One to six required |
| Product UX | Generate or edit in one model | Dedicated precision edit flow |
| Best starting point | New visual concept | Existing approved image |
Prompting Tips
- Put the subject and commercial purpose first
- Describe lighting and lens language instead of vague quality adjectives
- For edits, list preserved details before the requested change
- Use a platform-specific aspect ratio at generation time
- Change one major variable per edit so failures are diagnosable
Keep Model IDs Configurable
Store dreamina-v3 with the campaign template. You can compare the same brief
on GPT Image, Z-Image Turbo, or another model without changing authentication,
credits, or job polling.
API FAQ
Does Dreamina V3 support image editing?
Yes. Add one URL in imageUrls; the MakeInfluencer adapter selects the editing variant.
How large is the output?
The adapter computes valid dimensions for the requested aspect ratio with a longest edge targeted at 2048 pixels. Check the current model page for any contract changes.
How is the result returned?
Generation returns HTTP 202 and one or more image IDs. Poll /api/v1/generations/image/{generationId} until terminal is true. On success, ready is true and the response includes the image URLs.
Does the API use separate credits?
No. Dreamina API generations use the same MakeInfluencer balance as the dashboard and MCP connector.