Alibaba logo
fal-ai/qwen-image-2/pro/text-to-image

Qwen-Image-2.0 is a next-generation foundational unified generation-and-editing model
Inference
Commercial use
Partner

Input

Additional Settings

Customize your input with more control.

Result

Idle

What would you like to do next?

Your request will cost $0.075 per image.

Logs

Qwen Image 2.0 Pro — Text to Image

Qwen Image 2.0 Pro is the highest-fidelity text-to-image endpoint from Alibaba's Qwen team. It generates images natively at up to 2048 x 2048 pixels with professional typography rendering, detailed compositions, and support for infographics, posters, and comics. Ranked #1 on AI Arena for text-to-image generation.

When to Use Pro vs Standard

Use the Pro endpoint for final production assets where detail, text accuracy, and composition quality matter most. For rapid iteration and prototyping, the standard endpoint runs faster at lower cost ($0.035/image vs $0.075/image).

Key Parameters

ParameterDefaultRangeNotes
promptup to 1,000 tokensFront-load your main subject for best results
negative_promptstringElements to exclude (e.g. "blurry, watermark")
image_sizesquareenum or customsquare, square_hd, landscape_4_3, landscape_16_9, portrait_4_3, portrait_16_9
seedrandomintegerFix for reproducible results
num_images11–4Generate multiple variations per request
output_formatpngpng / jpeg / webpChoose based on your delivery needs

Quick Start

python
import fal_client

result = fal_client.subscribe(
    "fal-ai/qwen-image-2/pro/text-to-image",
    arguments={
        "prompt": "Minimalist movie poster for 'ECHO STATION', bold sans-serif title at top, lone astronaut in alien desert, two moons, muted teal and burnt orange palette",
        "image_size": "portrait_4_3"
    }
)

print(result["images"][0]["url"])
javascript
import { fal } from "@fal-ai/client";

const result = await fal.subscribe("fal-ai/qwen-image-2/pro/text-to-image", {
  input: {
    prompt: "Minimalist movie poster for 'ECHO STATION', bold sans-serif title at top, lone astronaut in alien desert, two moons, muted teal and burnt orange palette",
    image_size: "portrait_4_3"
  },
});

console.log(result.data.images[0].url);

Prompting Tips

  • Structure prompts as: Subject, style, details, composition, lighting. The model weights information by position, so put the most important elements first.
  • For text-heavy images (posters, infographics, slides): increase guidance_scale to 7–10 and num_inference_steps to 35+ for sharper text rendering.
  • Use negative prompts to suppress common artifacts: "blurry, low quality, distorted, deformed, oversaturated, watermark".
  • Specify aspect ratio to match your output context — landscape_16_9 for video thumbnails, portrait_4_3 for mobile, square_hd for maximum detail.

Best Use Cases

  • Infographics and data visualizations — the model can compose complex multi-section layouts with accurate text placement
  • Movie posters and print design — professional typography with correct perspective on surfaces
  • Product photography — editorial-quality renders with controlled lighting and composition
  • Comics and storyboards — multi-panel layouts with consistent character rendering
  • Social media assets — branded content with text overlays rendered directly in the image