Qwen-Image
Qwen-Image is Alibaba's Qwen team's image model, released on 4 August 2025 with open weights under Apache 2.0: a 20-billion-parameter diffusion transformer built for text rendering, including Chinese, with an editing sibling that followed two weeks later. It is the version you can download and run. Qwen-Image 2.0 was announced in February 2026 and Qwen-Image 3.0 in July 2026, both hosted only at launch, and 2.0's fal endpoint is already deprecated. On fal the original costs $0.02 a megapixel.
Anvisha Pai, Co-founder & CEO, Voyager
Alibaba · released
›What we asked for
A self-portrait of Qwen-Image, Alibaba's open-weight image model, as a bilingual shop sign: a night market stall with a hand-painted board reading 'Qwen-Image' in English and Chinese, lanterns, steam, a painter's brush resting on the ledge.
fal-ai/qwen-image
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1360 by 1024 pixels
At a glance
- Price
- $0.02 a megapixel on fal, rounded up; the eleven prompts that ran cost $0.38, the edit prompt having no endpoint.
- Open weights
- Apache 2.0, 20 billion parameters, on Hugging Face and ModelScope, with a 2512 refresh in December 2025.
- Text
- Built for typography, including Chinese, which is the reason to try it.
- Sizes
- Any size up to about 2K through fal's presets or custom pixels.
- Newer versions
- 2.0 and 3.0 are hosted-only and not on this site; 3.0 shipped without weights or a licence.
Other Qwen-Image models: Qwen Image 3. Compare all 2.
Anvisha's take
Qwen-Image earns its reputation on the poster and loses it on anything with small text. The three-line title card is bold and correct, the logo is a clean flat mark, the pour-over face is natural and the placement test is exact, all for a few cents at 1K. But the chart's month labels came out as 'Muy', 'The', 'Months' and 'Jnth', and the infographic's step wording dissolved into shapes on a photographed card. The self-portrait, a night-market painter under a bilingual sign, is the one hero on the site with correct Chinese in it. Open weights, cheap and fast; check every label.
Qwen-Image real output: the same 12 prompts every model gets
Every image model on this site gets the same 12 prompts, so you can compare like with like. These are the pictures Qwen-Image returned, one try each, nothing re-rolled or retouched. Generated on . The 11 pictures took 5 to 26 seconds each and cost $0.38 in total.
Spelling, kerning and hierarchy of three lines of set text.
›What we asked for
A concert poster that reads 'MODA SUMMER SESSIONS' in large bold type, with the subtitle 'Friday 12 June, 8pm, Roundhouse' and a smaller line 'Tickets at the door'. Risograph style, two colours, tangerine and navy, grainy ink texture.
fal-ai/qwen-image
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 1024 by 1360 pixels
Catalogue lighting, a clean seamless background and a believable contact shadow.
›What we asked for
Studio product photograph of a matte black ceramic pour-over coffee dripper on a seamless white background, soft top light, subtle contact shadow, centred, catalogue style.
fal-ai/qwen-image
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
A natural face, correct finger count and hands holding an object.
›What we asked for
Close-up portrait of a woman in her 40s with short grey hair, laughing, holding a red ceramic mug with both hands, natural window light, 85mm lens, shallow depth of field.
fal-ai/qwen-image
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 1024 by 1360 pixels
A flat single-colour mark with a legible wordmark and no gradients or photo texture.
›What we asked for
A flat vector logo mark for a bicycle repair shop called 'Spoke & Chain': a single-colour icon combining a chain link and a wheel spoke, with the wordmark below in a geometric sans-serif, black on white.
fal-ai/qwen-image
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Many distinct objects and people in one frame without merging or duplicating them.
›What we asked for
A crowded Saturday farmers market seen from slightly above: at least twelve stalls with striped awnings, shoppers carrying tote bags, a dog on a leash, crates of oranges, a musician with a guitar, bunting overhead, late morning light.
fal-ai/qwen-image
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1360 by 1024 pixels
A named historical style applied faithfully rather than a generic painterly look.
›What we asked for
A lighthouse on a rocky cliff during a storm, in the style of a Japanese ukiyo-e woodblock print, with visible wood grain texture, flat colour areas and a bokashi gradient sky.
fal-ai/qwen-image
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 1024 by 1360 pixels
Six labelled bars at the stated heights, in order, with readable axes.
›What we asked for
A clean bar chart titled 'Monthly active users' showing six months, January through June, with the values 120, 150, 180, 170, 220 and 260, labelled axes, white background, one accent colour, no 3D effects.
fal-ai/qwen-image
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1360 by 1024 pixels
A 21:9 composition that leaves the left third clear for headline text.
›What we asked for
Website hero banner: a wide panorama of a sunlit modern co-working space with plants, the left third of the frame kept as clear wall space for text, warm neutral palette.
fal-ai/qwen-image
image_size: {"width":2016,"height":864}, output_format: png, num_images: 1
returned 1536 by 864 pixels
Four numbered steps in the given order with their exact wording, plus an icon each.
›What we asked for
An infographic card titled 'How to brew pour-over' with four numbered steps in this order: 1. Rinse the filter, 2. Bloom for 30 seconds, 3. Pour in slow circles, 4. Wait 3 minutes. A simple icon beside each step, muted palette, clean sans-serif type.
fal-ai/qwen-image
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 1024 by 1360 pixels
Appetising, menu-grade food with correct textures and no melted anatomy.
›What we asked for
Overhead photograph of a margherita pizza on a wooden board, fresh basil leaves, one slice pulled out with the cheese stretching, restaurant menu photography, natural light.
fal-ai/qwen-image
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Exact counts and left-right placement of five named objects.
›What we asked for
A glass of water on a wooden table with exactly three ice cubes in it, a yellow pencil resting across the rim of the glass, a blue book to the left of the glass and a red apple to the right, photographed from table height.
fal-ai/qwen-image
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1360 by 1024 pixels
Change one object's colour and add a part while leaving background, light and framing untouched.
›What we asked for
Change the coffee dripper to a deep forest green and add a small brass handle on its right side. Keep the background, lighting and composition exactly as they are. (applied to the product on white result above)
fal-ai/qwen-image
How the cost under each picture was worked out: fal's Qwen-Image rate of $0.02 per megapixel (fal model page, read 2026-09-10), rounded up to the nearest 1024x1024 megapixel as fal states: 2 for the 1024x1360 output; fal does not return the charge.
Qwen-Image pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| fal | Megapixel Rounded up to the nearest megapixel; a 1024 by 1024 image is one. | $0.02 | fal model page, read |
| Self-hosted | Image Free under Apache 2.0; you pay for the GPU. | $0.00 | QwenLM/Qwen-Image on GitHub, read |
Is Qwen-Image free?
The weights are free under Apache 2.0, and Qwen Chat generates images free with limits. Hosted endpoints charge per megapixel.
- •Open weights: Qwen-Image, Qwen-Image-Edit and the 2512 refresh on Hugging Face and ModelScope, Apache 2.0.
- •Qwen Chat: free image generation with daily limits.
- •fal and other hosts: per megapixel, no free tier.
Where to use Qwen-Image
- Qwen Chat · Alibaba's chat app with free image generation.
- Open weights · Apache 2.0 weights and inference code.
- fal · $0.02 a megapixel; a separate edit endpoint exists.
- Voyager · Runs inside Voyager, the agent for creative work. Private preview by waitlist.
Voyager
Run Qwen-Image inside Voyager
Voyager is an agent for creative work that runs every model in this reference, Qwen-Image included. Brief it, and it picks the model, makes the thing and hands back something you can edit. Create an account to get started.
What we noticed running Qwen-Image
- •The bars and values are right; the month labels read 'Muy', 'January', 'The', 'Months', 'Jnth', 'June'. Small axis text is where it breaks. See the chart picture.
- •Returned a photograph of a printed card with 'pour-oyf' in the title and the step wording turned to marks; the four icons survived. See the infographic with steps picture.
- •Bold, correct, three lines, the layout of a real gig poster. See the text in an image picture.
- •Self-portrait: the only hero on the site with readable Chinese characters, on a sign that also spells the model's name in English.
Qwen-Image limits and API parameters
| Limit | Value |
|---|---|
| Sizes | Presets or custom pixels; our runs used 1024 by 1024 up to 1536 by 864. fal API reference, read |
| Editing | Not on this page: the edit sibling is a separate endpoint we did not run. fal API reference, read |
›Every setting, for developers
fal (fal-ai/qwen-image)
fal API reference, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | Any text Required. | |
| image_size | enum or {width, height} | square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3, landscape_16_9, or custom pixels | landscape_4_3 |
| num_inference_steps | integer | Steps | 30 |
| guidance_scale | number | Prompt adherence | 2.5 |
| acceleration | enum | none, regular, high | none |
| negative_prompt | string | Any text | |
| num_images | integer | 1 and up | 1 |
| output_format | enum | jpeg, png | png |
| seed | integer | Any integer |
What Qwen-Image will not draw
The hosted endpoint runs behind fal's safety checker; the open weights carry Apache 2.0 and Alibaba's acceptable-use terms. Self-hosted, the model has no filter beyond what you add.
fal API reference, read
Alternatives to Qwen-Image
Frequently asked questions
What is Qwen-Image?
Qwen-Image is Alibaba's open-weight image model from 4 August 2025, a 20-billion-parameter diffusion transformer under Apache 2.0, built for text rendering including Chinese. It is on fal at $0.02 a megapixel.
Is Qwen-Image open source?
The original and its December 2025 refresh are open weights under Apache 2.0. Qwen-Image 2.0 and 3.0 were hosted-only at launch; 3.0 shipped without weights or a licence.
Is Qwen-Image free?
The weights are free to run yourself, and Qwen Chat generates images free with limits. Hosts charge per megapixel.
How this page is made
Prices, limits and settings are read from the linked vendor and host pages on the dates shown. Every picture was generated by us through fal with the request shown under it, and the original files are kept unedited. A generation that failed is shown as a failure. Model and vendor names are used to identify the products; no vendor imagery appears on this page.
