Cookie Preferences
We use cookies to personalize content and analyze traffic. See our Cookie Policy.
No image yet. Submit the form to generate image.
Explore different use cases and parameter configurations
Transparent pricing with no hidden fees. Pay as you go.
| Rule & Modality | Channel | Credits | Price (USD) | Official / Reference Price | Daily Savings |
|---|---|---|---|---|---|
Wan-2.7-image-pro-1K imagewanresolution: 1K | Starter | 1per run | $0.005 | - | - |
Wan-2.7-image-pro-2K imagewanresolution: 2K | Starter | 1per run | $0.005 | - | - |
Wan-2.7-image-pro-4K imagewanresolution: 4K | Starter | 1per run | $0.005 | - | - |
Complete guide to using wan 2.7 image
Generate production-ready 4K visuals with precise control and advanced chain-of-thought reasoning.
Wan 2.7 Image Generation Pro is Alibaba's flagship professional-grade AI image generation model, purposefully built for developers and creators who demand high-fidelity, production-ready visuals. Unlike conventional diffusion models, Wan 2.7 Pro utilizes a unique chain-of-thought reasoning mechanism—meaning the model literally "thinks before it draws" by analyzing composition logic, lighting semantics, and layout relationships before rendering pixels. This API sets itself apart from the standard Wan 2.7 Image model by delivering massive 4K resolution output, providing more stable composition, stronger visual consistency, and achieving sharper prompt understanding to meet the rigorous demands of professional creative workflows .
Submit a prompt without input_urls to generate fresh images. In this mode, aspect_ratio can be forwarded, and thinking_mode may be enabled when enable_sequential is false.
Provide one or more images through input_urls to edit existing visuals. You can optionally attach bbox_list to describe local edit regions per input image.
Enable enable_sequential to request grouped image generation. In this mode, the adapter allows n up to 12 and intentionally disables features that the upstream model restricts, such as thinking_mode and color_palette.
Wan 2.7 Image Pro API outputs incredibly detailed, print-ready images up to 4K (4096×4096 pixels) . This makes it the superior choice for high-DPI deliverables, large-format displays, and professional print workflows .
Wan 2.7 Image Pro API features an enhanced built-in reasoning mode that evaluates scene structure and spatial logic before generation. This results in noticeably sharper prompt understanding, tighter composition, and significantly fewer AI hallucinations compared to the standard model .
Designed for demanding projects, the Pro API offers stronger consistency than the standard version by allowing you to input up to 9 reference images simultaneously. This ensures strict style locking, exact background alignment, and perfect character identity preservation across entire series of visual assets.
Wan 2.7 Image Pro API mpower your text-heavy visuals with the ability to render highly legible typography, academic formulas, and structured tables directly into your 4K images. The Pro API reliably processes up to 3,000 tokens of text input across 12 different languages.
Go beyond basic generation with advanced instruction-based image editing capabilities that allow for pixel-level accuracy. The Pro version provides advanced professional control, allowing you to fine-tune facial features, extract palettes, and enforce precise brand alignment using exact HEX color codes.
The integration flow follows the same asynchronous jobs API used by the other Kie-backed models.
Choose wan/wan-2-7-image-pro as the model and prepare the input object with prompt plus any optional Wan-specific parameters such as input_urls, aspect_ratio, n, resolution, bbox_list, seed, and nsfw_checker.
Send the payload to the ApiPass /api/v1/jobs/createTask endpoint. The adapter maps the public model name to the upstream Kie.ai model and returns a taskId when the request is accepted.
Track the task through the query endpoint or your callback URL. The adapter maps provider task states into ApiPass states such as queuing, generating, success, and fail.
After the task succeeds, consume resultUrls from the task result object. For completed tasks, the adapter attempts to upload the upstream files to the configured CDN and returns the CDN-backed URLs when available.
These suggestions are based on the request constraints implemented by the adapter and documented by the provider.
If you want to control the output composition or preserve source content, provide input_urls. If you are creating fresh images from text, leave input_urls empty so aspect_ratio and thinking_mode can be used normally.
bbox_list should use the same outer-array length as input_urls. The adapter pads or trims mismatched lengths, but supplying aligned data avoids ambiguity around which boxes belong to which image.
Enable sequential mode only when you need grouped outputs or more than four images. It changes the valid range for n and disables some optional controls such as thinking_mode and color_palette.
Use input images plus bbox_list to replace ingredients, props, labels, or surface details while preserving the original camera angle, lighting, and scene structure.
Generate multiple still-image options from a single prompt by adjusting n, resolution, and aspect_ratio for landing pages, ad mockups, and product campaigns.
Enable sequential mode to produce grouped variants such as character turnarounds, costume studies, or style exploration sets from one prompt.
The document uses wan/wan-2-7-image-pro as the public ApiPass model name because that is the current value in SUPPORT_MODEL and the one expected by the platform.
The readme and API sections explain that ApiPass forwards the public model to the upstream Kie.ai identifier wan/2-7-image-pro, matching the adapter mapping implemented in _get_kie_api_model_name.
The create-task API section documents the public ApiPass payload shape of model plus optional callBackUrl and input. That is the same shape the adapter receives before forwarding the task to Kie.ai.
All APIs require authentication via Bearer Token.
Authorization: Bearer
Create a new Wan 2.7 Image Pro generation or editing task through the ApiPass unified jobs API
ApiPass keeps the public request envelope stable across async image models. Wan-specific parameters are placed inside the input object.
1{
2 "model": "wan/wan-2-7-image-pro",
3 "callBackUrl": "string (optional)",
4 "input": {
5 "prompt": "string",
6 "input_urls": [
7 "string"
8 ],
9 "aspect_ratio": "1:1 | 16:9 | 4:3 | 21:9 | 3:4 | 9:16 | 8:1 | 1:8",
10 "enable_sequential": false,
11 "n": 4,
12 "resolution": "1K | 2K | 4K",
13 "thinking_mode": false,
14 "color_palette": [
15 {
16 "hex": "#C2D1E6",
17 "ratio": "23.51%"
18 }
19 ],
20 "bbox_list": [
21 []
22 ],
23 "watermark": false,
24 "seed": 0,
25 "nsfw_checker": false
26 },
27 "channel": "auto"
28}modelRequiredstringThe public ApiPass model name for Wan 2.7 Image Pro.
"wan/wan-2-7-image-pro"
callBackUrlOptionalstringOptional callback URL for task completion notifications.
"https://your-domain.com/api/callback"
channelOptionalstringYou may specify the corresponding provider within APIPASS via the channel parameter; these providers handle the actual image and video generation tasks. APIPASS currently offers three provider options:
The default value for the channel parameter is auto. When enabled, APIPASS automatically allocates tasks across available providers based on real-time pricing and stability metrics to balance minimal cost and reliable performance. Retain the default auto value unless you have custom routing requirements.
Available options:
auto
The input object contains the Wan 2.7 Image Pro parameters accepted by ApiPass. The adapter reshapes and normalizes them before forwarding the request upstream.
input.promptRequiredstringPrompt used for generation or editing.
"A cozy reading corner with warm wood shelves and soft afternoon sunlight."
input.input_urlsOptionalarray(URL)Optional source images for editing workflows. The adapter also accepts image_urls, images, and image_input and converts them to input_urls.
["https://example.com/source.jpg"]
input.aspect_ratioOptionalstringOutput aspect ratio. This is only forwarded when input_urls is empty.
Available options:
"16:9"
input.enable_sequentialOptionalbooleanWhether grouped output mode is enabled. This changes the valid range for n.
false
input.nOptionalnumberNumber of requested outputs. ApiPass clamps this to 1-4 when enable_sequential is false and 1-12 when it is true.
4
input.resolutionOptionalstringRequested output resolution. Unsupported values are normalized to 2K.
Available options:
"2K"
input.thinking_modeOptionalbooleanOptional thinking mode. The adapter only enables it when input_urls is empty and enable_sequential is false.
true
input.color_paletteOptionalarray(object)Optional color theme definitions. This field is forwarded only when enable_sequential is false and the array contains 3-10 entries.
[{"hex":"#D7C2A3","ratio":"32.00%"}]input.bbox_listOptionalarray(array)Optional editing boxes aligned with input_urls. Each image can include up to 2 boxes, where each box is [x1, y1, x2, y2].
[[120,80,320,260]]
input.watermarkOptionalbooleanWhether to add a watermark to generated images.
false
input.seedOptionalnumberOptional seed value. Invalid values are normalized to 0 by the adapter.
0
input.nsfw_checkerOptionalbooleanOptional NSFW checker flag. The adapter also accepts nsfwChecker and normalizes boolean-like input values.
false
1curl -X POST "https://api.apipass.dev/api/v1/jobs/createTask" \
2 -H "Content-Type: application/json" \
3 -H "Authorization: Bearer YOUR_API_KEY" \
4 -d '{
5 "model": "wan/wan-2-7-image-pro",
6 "callBackUrl": "https://your-domain.com/api/callback",
7 "input": {
8 "prompt": "Transform the food photo by replacing the marked ingredients with sliced red chili pieces in the corresponding positions, while keeping the same bowl, composition, steam, lighting, camera angle, background, texture, and overall realism.",
9 "input_urls": [
10 "https://static.aiquickdraw.com/tools/example/1775122744247_eSHwJX1k.jpg"
11 ],
12 "n": 1,
13 "enable_sequential": false,
14 "resolution": "2K",
15 "thinking_mode": false,
16 "watermark": false,
17 "seed": 0,
18 "nsfw_checker": false,
19 "bbox_list": [
20 []
21 ]
22 }
23 }'1{
2 "code": 200,
3 "message": "success",
4 "data": {
5 "taskId": "task_12345678"
6 }
7}codeStatus code. 200 indicates the task was created successfully.
messageResponse message returned by ApiPass.
data.taskIdTask identifier used for later status checks.
flux
flux
Generate photorealistic images with robust text
Starting from
0 credits
Nano Banana 2 Lite, aka Gemini 3.1 Flash-Lite Image, is a newly released image model from Google DeepMind built for fast 1K image generation, efficient creative iteration, prompt-based image editing, and responsive visual output in product-facing creative scenarios.
Starting from
0 credits
OpenAI
OpenAI
Generate intentionally designed visuals with less prompting using the latest OpenAI GPT Image 2 model.
Starting from
0 credits
Nano Banana now allows you to freely draw on edited images, bringing you a completely new experience. Simply click on the image to draw whatever you want. You can also add text and details, and Gemini will do the rest.
Starting from
0 credits
Fast image generation model built on Gemini 3.1 Flash Image. Supports text-to-image, image editing, multi-image fusion with up to 14 reference images, Google Search grounding, and outputs up to 4K resolution.
Starting from
0 credits
seedream
seedream
Starting from
0 credits
grok
grok
Grok Imagine is xAI’s AI video generation model for creating high-quality videos with realistic motion, native audio, and synchronized speech. Grok Imagine 1.5 introduces faster generation speeds, improved motion physics, and stronger visual consistency for modern AI video applications.
Starting from
0 credits
qwen
qwen
Generate realistic 2K images with structured text
Starting from
0 credits
seedream
seedream
Starting from
0 credits
OpenAI
OpenAI
Bring your ideas to life with sharper, faster, more precise image generation using the GPT Image 2.5 API.
Starting from
0 credits
seedream
seedream
Gemini said Generate images with lightning-fast efficiency and high-fidelity detail through a streamlined, lightweight architecture optimized for instant creative expression.
Starting from
0 creditsqwen
qwen
Generate realistic 2K images with structured text
Starting from
0 credits
Luma
Luma
Starting from
0 credits
seedream
seedream
Starting from
0 credits