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 |
|---|---|---|---|---|---|
qwen-image-2 imageqwen | Starter | 0.7per run | $0.003 | - | - |
Complete guide to using qwen image 2
Generate still images through the ApiPass Kie adapter using the public model name qwen/qwen-image-2.
Available now on APIPASS, Alibaba's Qwen Image 2.0 is a highly versatile, unified AI model designed to handle realistic image generation, structured text rendering, and flexible natural-language image editing. Engineered to support comprehensive creative and visual design workflows, it delivers stunning native 2K output.
Every request requires a prompt. The adapter rejects empty prompt payloads before any upstream request is attempted.
The model supports square, square_hd, portrait_4_3, portrait_16_9, landscape_4_3, and landscape_16_9. The adapter also accepts imageSize and normalizes unsupported values to square_hd.
You can adjust num_inference_steps, guidance_scale, enable_safety_checker, output_format, negative_prompt, acceleration, and seed. The adapter clamps or normalizes unsupported values before forwarding the request upstream.
On APIPASS, the Qwen Image 2.0 API effortlessly processes detailed prompt instructions to create text-heavy visuals. Whether you are designing PPT-style slides, complex infographics, comics, or posters, this API ensures highly readable in-image text and precise layout control, making it the perfect engine for design-focused visual generation.
Experience unmatched precision when bringing your ideas to life. The Qwen Image 2 API delivers native 2K support and exceptional semantic adherence, meaning it accurately interprets complex visual instructions. The result is highly detailed, polished, and realistic imagery across subjects like people, nature, and complex architectural scenes.
Say goodbye to disjointed creative processes. APIPASS offers Qwen-Image 2.0 as a unified solution that combines prompt-based generation and follow-up visual editing within the exact same model workflow. This provides a highly consistent and efficient path for products requiring both initial image creation and subsequent refinements.
Speed up your visual tasks with the Qwen Image Edit API, built on a highly efficient, lightweight 7B-parameter model architecture. Designed for rapid response times, it easily handles practical editing commands in natural language—allowing you to instantly execute background swaps, object updates, style adjustments, and text revisions directly through APIPASS.
The integration flow matches the other Kie-backed asynchronous image models on ApiPass.
Choose qwen/qwen-image-2 as the model and prepare the input object with prompt plus any optional Qwen controls such as image_size, guidance_scale, output_format, and negative_prompt.
Send the payload to the ApiPass /api/v1/jobs/createTask endpoint. The adapter rewrites the model name to qwen/text-to-image and forwards the normalized input object to Kie.ai.
Track task execution through /api/v1/jobs/recordInfo or configure callBackUrl to receive asynchronous completion notifications.
After the task succeeds, read the generated image URLs from resultJson.resultUrls in the task response object.
These recommendations follow the adapter guardrails and the current upstream parameter contract.
Select portrait_16_9 for tall layouts, landscape_16_9 for banners, and square_hd for general-purpose artwork or marketplaces before refining prompt wording.
Qwen image quality tends to improve when you explicitly exclude common defects such as blur, extra fingers, low contrast, distorted anatomy, or noisy textures.
Start with moderate num_inference_steps and acceleration set to none or regular for exploration. Increase steps only when you need more fidelity after the composition is already working.
Create banner-ready or social-ready still images by pairing strong prompt composition with the right image_size preset and optional negative_prompt constraints.
Rapidly iterate on environments, products, and visual directions using seed values, size presets, and controlled prompt variation.
Generate clean, controlled compositions for landing pages, explainer sections, blog covers, and feature showcases while explicitly suppressing common visual artifacts.
The document uses qwen/qwen-image-2 as the public ApiPass model name because that is the current SUPPORT_MODEL value used by the codebase and example script.
The readme and API sections explain that ApiPass forwards qwen/qwen-image-2 to the upstream Kie.ai model qwen/text-to-image, matching the adapter mapping implemented in _get_kie_api_model_name.
The JSON documents the same coercion rules implemented in the adapter for image_size, num_inference_steps, guidance_scale, output_format, acceleration, and compatibility aliases such as imageSize and enableSafetyChecker.
All APIs require authentication via Bearer Token.
Authorization: Bearer
Create a new Qwen Image 2 generation task through the ApiPass unified jobs API
ApiPass keeps the public request envelope stable across asynchronous Kie-backed image models. Qwen-specific parameters are placed inside the input object.
1{
2 "model": "qwen/qwen-image-2",
3 "callBackUrl": "string (optional)",
4 "input": {
5 "prompt": "string",
6 "image_size": "square | square_hd | portrait_4_3 | portrait_16_9 | landscape_4_3 | landscape_16_9",
7 "num_inference_steps": 30,
8 "seed": 123456,
9 "guidance_scale": 2.5,
10 "enable_safety_checker": true,
11 "output_format": "png | jpeg",
12 "negative_prompt": "string",
13 "acceleration": "none | regular | high"
14 },
15 "channel": "auto"
16}modelRequiredstringThe public ApiPass model name for Qwen Image 2.
"qwen/qwen-image-2"
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 Qwen Image 2 parameters accepted by ApiPass. The adapter normalizes aliases and guards invalid values before forwarding the request upstream.
input.promptRequiredstringPrompt describing the image you want to generate.
"A cinematic mountain lake at sunrise with mist and reflective water."
input.image_sizeOptionalstringImage size preset. The adapter also accepts imageSize and normalizes unsupported values to square_hd.
Available options:
"portrait_16_9"
input.num_inference_stepsOptionalnumberInference step count. The adapter also accepts numInferenceSteps and clamps this value to 2-250.
40
input.seedOptionalnumberOptional seed. The adapter attempts to coerce the incoming value to an integer and skips invalid values.
123456
input.guidance_scaleOptionalnumberGuidance scale. The adapter also accepts guidanceScale and clamps this value to 0-20.
3.25
input.enable_safety_checkerOptionalbooleanWhether to enable the provider safety checker. The adapter also accepts enableSafetyChecker and defaults this field to true.
true
input.output_formatOptionalstringImage output format. The adapter accepts jpg and converts it to jpeg, while unsupported values fall back to png.
Available options:
"jpeg"
input.negative_promptOptionalstringOptional negative prompt. The adapter also accepts negativePrompt and defaults it to a single space when omitted.
"blurry, low quality"
input.accelerationOptionalstringAcceleration mode. Unsupported values are normalized to none.
Available options:
"regular"
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": "qwen/qwen-image-2",
6 "callBackUrl": "https://your-domain.com/api/callback",
7 "input": {
8 "prompt": "A cinematic astronaut exploring a crystal cave on an alien planet, volumetric lighting, ultra detailed.",
9 "image_size": "square_hd",
10 "num_inference_steps": 30,
11 "seed": 123456,
12 "guidance_scale": 2.5,
13 "enable_safety_checker": true,
14 "output_format": "png",
15 "negative_prompt": "blurry, low resolution, distorted hands, extra fingers",
16 "acceleration": "none"
17 }
18 }'1{
2 "code": 200,
3 "message": "success",
4 "data": {
5 "taskId": "task_qwen_1765166238716"
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 creditsqwen
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
wan
wan
Generate images with cinematic precision through high-fidelity text-to-image synthesis and structural editing.
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 credits
Luma
Luma
Starting from
0 credits
seedream
seedream
Starting from
0 credits