Alibaba's most advanced image generation and editing model. Understands instructions up to 4,500 tokens in a single pass, renders crisp text in 12 languages and 20+ fonts, and delivers hyperrealistic skin, hair, and material detail at up to 2K resolution.
Qwen Image 3 renders crisp, readable text directly into images, down to roughly 10 pixels. Native support for 12 languages and more than 20 fonts makes posters, slide decks, product packaging, comics, and long-form infographics achievable in a single generation, with text that adapts to surfaces and perspective.
Instructions of up to 4,500 tokens let you specify layout, exact text content, style, and fine detail all at once. Combined with strong world knowledge, the model composes nested UI mockups, multi-panel storyboards, and knowledge-rich infographics in a single pass, no model chaining required.
Skin, hair, fabric, and paper are reproduced with close-to-photographic fidelity at up to 2048 × 2048 resolution. The result is production-ready imagery built for user-generated content, portraits, and product photography, holding up at full resolution without post-processing.
Generate from a prompt or edit an existing image with the same model. Both endpoints are pay-per-image at $0.075.
Generates images from a text prompt at resolutions up to 2048×2048, with automatic prompt rewriting and prompt-guided resolution selection, building on Qwen's strength in complex text rendering and precise prompt adherence
Edits images from one to three reference images and a natural-language instruction, preserving key details such as facial features and identity while applying the requested changes
A selection of images generated with Qwen Image 3 showcasing infographics, multilingual typography, hyperrealistic portraits, and product photography.
"A four-panel educational comic in Chinese explaining the water cycle, cheerful animated water-drop characters with speech bubbles, legible captions for evaporation, condensation, precipitation, and collection, bright flat illustration style"
"Hyperrealistic editorial portrait of a young woman in warm afternoon light, dappled floral shadows across her face, holding a flower, natural skin texture and freckles, shallow depth of field, magazine beauty photography"
"A labeled science diagram comparing a plant cell and an animal cell side by side, detailed organelle illustrations with annotation labels and a key-differences table, clean textbook layout, crisp readable English text"
"Professional product photograph of a frosted glass perfume bottle on a wet slate surface, brand name 'AURELIA' etched into the glass, soft studio lighting with a single reflection, deep green and gold accents, luxury advertising composition"
The client API handles the request submit protocol. It will handle the request status updates and return the result when the request is completed.
import { fal } from "@fal-ai/client";
const result = await fal.subscribe("alibaba/qwen-image-3/text-to-image", {
input: {
prompt: "A trilingual coffee shop poster reading 'MORNING BLEND' in English, Chinese, and Japanese, warm minimalist layout, a steaming ceramic cup at the center",
image_size: "portrait_4_3",
num_images: 1,
},
logs: true,
onQueueUpdate: (update) => {
if (update.status === "IN_PROGRESS") {
update.logs.map((log) => log.message).forEach(console.log);
}
},
});
console.log(result.data);
console.log(result.requestId);Qwen Image 3 is Alibaba's latest image foundation model from the Qwen team, released in July 2026. It unifies text-to-image generation and image editing in one model. It understands prompts up to 4,500 tokens, renders legible text in 12 languages across more than 20 fonts, and produces hyperrealistic detail at up to 2K resolution.
Qwen Image 3 renders crisp, readable text directly in generated images, down to roughly 10 pixels. It supports 12 languages and more than 20 fonts, so a single generation can produce multilingual posters, slide decks, product packaging, nested UI layouts, comics, and long-form infographics. Text adapts to different surfaces with correct perspective and material properties.
The model accepts instructions of up to 4,500 tokens, so you can specify layout, exact text content, style, and fine detail in a single request. Prompt expansion is enabled by default to enrich shorter prompts, and you can toggle it off for full manual control.
Qwen Image 3 generates at up to 2048 × 2048 pixels (2K). The edit endpoint also supports prompt-guided resolution selection, choosing an appropriate output size from your instruction. Fine details like skin pores, hair, fabric weave, and paper texture are rendered with high fidelity during generation.
The edit endpoint takes one to three reference images plus a natural-language instruction. It preserves identity and facial features while applying requested changes, and includes automatic prompt rewriting for cleaner results. Qwen's strength in complex text rendering and precise prompt adherence carries into editing, making it well suited to adding or changing text within an existing image.
Pricing is pay-per-image with no minimums or subscriptions. Both text-to-image generation and image editing cost $0.075 per generated image. You only pay for the images you create.
Install the fal.ai SDK (Python or JavaScript), grab an API key from your dashboard, and make your first request in a few lines of code. The API is serverless, so there are no GPUs to manage and no infrastructure to set up. Check the API documentation for all available parameters.
Yes. Images generated through the fal.ai API can be used in commercial projects. Check fal.ai's terms of service for full details on usage rights and licensing.
Want to learn more about integrating Qwen Image 3 into your workflow? Leave your details and our team will reach out.
