Skip to main content
POST
/
v1
/
tasks
Wan Image Series
curl --request POST \
  --url https://www.qingbo.dev/v1/tasks \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "<string>",
  "action": "<string>",
  "prompt": "<string>",
  "n": 123,
  "seed": 123,
  "aspect_ratio": "<string>",
  "resolution": "<string>",
  "image_urls": [
    "<string>"
  ],
  "callback_url": "<string>",
  "callback_events": [
    "<string>"
  ],
  "watermark": true,
  "thinking_mode": true,
  "enable_sequential": true,
  "color_palette": [
    {}
  ],
  "bbox_list": [
    {}
  ]
}
'
{
  "task_id": "task-wave1775290140a830128812",
  "model": "wan2.7-image",
  "action": "generate",
  "status": "queued",
  "created_at": 1775290140120,
  "progress": 0
}

Documentation Index

Fetch the complete documentation index at: https://docs.qingbo.dev/llms.txt

Use this file to discover all available pages before exploring further.

Alibaba Tongyi Wanxiang 2.7 unified interface for image generation and editing. Covers wan2.7-image-pro (professional, text-to-image up to 4K) and wan2.7-image (standard, up to 2K) tiers. The two tiers share the same interface via internal tier switching, and are exposed as a single model externally. Six call modes:
  • Text-to-image (generate) — prompt only, with optional thinking mode
  • Image-to-image (image2image) — Free-form generation from a single reference
  • Image edit (edit) — Inpainting / style transfer / element replacement
  • Multi-image reference (reference) — Fuse up to 9 reference images
  • Image group (group) — enable_sequential for comic strips / storyboards, max n = 12
  • Interactive editing (interactive_edit) — bbox_list targeted region editing (exclusive)
Per-image pricing, unified at $0.02295 per image.

​
Pricing

ModelPrice (per image)
wan2.7-image$0.02295

​
Mode Quick Reference

ModeTrigger Fieldaction
Text-to-imageprompt only (optional thinking_mode)generate
Image-to-image+ image_urls (single)image2image
Image edit+ image_urls (single) + edit-instruction promptedit
Multi-image reference fusion+ image_urls (2-9)reference
Image group (comic / storyboard)enable_sequential: true + n ≤ 12group
Interactive bbox editing+ image_urls + bbox_listinteractive_edit

​
Examples

curl -X POST https://www.qingbo.dev/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "wan2.7-image",
    "prompt": "Classical Chinese landscape scroll, ink-wash mood, distant mountains in dark green, near water in mist",
    "resolution": "4K",
    "aspect_ratio": "16:9",
    "thinking_mode": true
  }'

{
  "task_id": "task-wave1775290140a830128812",
  "model": "wan2.7-image",
  "action": "generate",
  "status": "queued",
  "created_at": 1775290140120,
  "progress": 0
}
After submission, poll status with GET /v1/tasks/{task_id}. See Task System for details.

​
Available Models

Model IDDescription
wan2.7-imageAlibaba Tongyi Wanxiang 2.7. Unified interface for text-to-image / image-to-image / editing / multi-image reference / image groups / interactive bbox editing. Up to 4K

​
Common Parameters

​
model
string
required
Choose one from the Available Models list. Currently wan2.7-image
​
action
string
default:"generate"
Operation type. Options:
  • generate — Text-to-image (default)
  • image2image — Image-to-image (use with single image_urls)
  • edit — Image editing (inpainting / style transfer)
  • reference — Multi-image reference fusion (up to 9)
  • group — Image group generation (use with enable_sequential)
  • interactive_edit — Interactive bbox editing (use with bbox_list)
​
prompt
string
required
Image description. Supports Chinese and English. For image-group mode, storyboard-style descriptions work best
​
n
integer
default:"1"
Number of images. Standard mode is bounded by service-side limits; image-group mode (enable_sequential: true) allows up to 12
​
seed
integer
default:"-1"
Random seed. -1 for random; a fixed value reproduces similar results
​
aspect_ratio
string
default:"1:1"
Aspect ratio, 7 options:
  • 1:1 — Square
  • 16:9 / 9:16 — Widescreen landscape / portrait
  • 4:3 / 3:4 — Standard landscape / portrait
  • 3:2 / 2:3 — Camera ratio landscape / portrait
​
resolution
string
default:"2K"
Output resolution. Options: 1K / 2K / 4K. Note: image-group mode caps at 2K
​
image_urls
string[]
Array of reference image URLs:
  • image2image / edit / interactive_edit: 1 image
  • reference: 2-9 images
​
callback_url
string
Webhook callback URL, invoked when the task reaches a terminal state. See Callback Mechanism
​
callback_events
string[]
Subscribed callback event types. See Callback Mechanism

​
Model-Specific Parameters

​
watermark
boolean
default:"false"
AI-generated watermark
​
thinking_mode
boolean
default:"true"
Thinking mode — when enabled, the model performs visual reasoning before generation, improving parsing accuracy on complex prompts.Activation conditions: only applies when not in image-group mode (enable_sequential not true) and without image input (image_urls not provided); ignored otherwise
​
enable_sequential
boolean
default:"false"
Image-group mode (comic strip / storyboard). When enabled:
  • n cap raises to 12
  • resolution capped at 2K
  • thinking_mode and color_palette have no effect
Suited for storyboards, comic strips, and themed e-commerce sets — any “themed group of related images” use case
​
color_palette
array
Custom color palette. 3-10 items of {hex, ratio}, with ratio summing to 100%.Activation condition: only applies in non-group mode. See Composite Field Reference
​
bbox_list
array
Interactive editing bbox list, one-to-one with image_urls, max 2 boxes per image.Only applies when action: "interactive_edit". See Composite Field Reference

​
Composite Field Reference

​
color_palette — Custom Color Palette

Specifies primary colors and their proportions for the output. The model distributes color coverage in the image based on the given ratios. Structure: array, each item is {hex, ratio}
FieldTypeDescription
hexstringHex color, e.g. #0A2540, case-insensitive
rationumberPercentage coverage in the image. All items must sum to 100
Constraints:
  • Item count: 3-10
  • Sum of ratio: strictly 100 (integer percentages)
  • Only effective in non-group mode (ignored when enable_sequential is true)
Example: 
"color_palette": [
  {"hex": "#0A2540", "ratio": 50},
  {"hex": "#F5F5F0", "ratio": 35},
  {"hex": "#E04B3A", "ratio": 15}
]

​
bbox_list — Interactive Editing Boxes

Draw boxes on a reference image to tell the model “edit only inside the box,” leaving the outside unchanged. Exclusive capability of the Wan series for interactive editing. Structure: triple-nested array [[[x1, y1, x2, y2], ...], ...]
  • Outermost: each item corresponds to one image in image_urls, one-to-one (counts must match)
  • Middle: multiple boxes on a single image, max 2 boxes per image
  • Innermost: coordinates of one box, 4 numbers [x1, y1, x2, y2], top-left and bottom-right corners (pixel coordinates)
Constraints:
  • Only applies when action: "interactive_edit"
  • Max 2 boxes per image
  • Coordinates use the reference image’s original pixels
Example (single image, single box): 
"image_urls": ["https://cdn.example.com/cat.jpg"],
"bbox_list": [[[120, 80, 540, 460]]]
Example (single image, two boxes): 
"image_urls": ["https://cdn.example.com/scene.jpg"],
"bbox_list": [[[100, 100, 400, 400], [500, 200, 800, 500]]]
Example (multiple images, one box each): 
"image_urls": [
  "https://cdn.example.com/img-a.jpg",
  "https://cdn.example.com/img-b.jpg"
],
"bbox_list": [
  [[120, 80, 540, 460]],
  [[200, 150, 600, 500]]
]

​
Resource Limits

ItemLimit
Reference images (reference)2-9
Reference images (image2image / edit / interactive_edit)1
Single image size≤ 30MB, supports JPG/PNG/WEBP
Output resolution1K / 2K / 4K (group mode caps at 2K)
Group count nUp to 12 (requires enable_sequential: true)
bbox_list boxesMax 2 per image
color_palette items3-10, ratio sums to 100%
Output fileJPG URL, valid for 24 hours