Nano Banana

Qwen Image 2.1 — Open Image Generation and Editing in One Model

Qwen-Image-2.1, released by Alibaba's Qwen team in September 2026, is a 7B open model that generates and edits images from a single checkpoint. It improves text and portrait aesthetics and became the #1 open-source model in both the Image Edit and Text-to-Image Arenas.

  • Alibaba's Qwen-Image-2.1
  • #1 open model in Image Edit and T2I Arenas
  • Edit with up to 10 references
  • 1K and 2K
Open the full workspace

More models

Why Qwen Image 2.1 Is a Great AI Image Generator

Qwen Image 2.1 is an affordable all-rounder: generate from scratch, then edit the same image with precise local instructions — all with one model.

  • Generate and edit

    One checkpoint handles both text-to-image and image editing.

  • Better text and faces

    Qwen improved the look of rendered text and portraits over 2.0.

  • Many references

    Combine up to 10 reference images in one edit.

  • Open-source leader

    Ranked #1 among open models in both the Image Edit and Text-to-Image Arenas.

What Qwen Image 2.1 can do

  • Text to image

    Posters, portraits and scenes from a prompt.

  • Local edits

    Describe which region to change and what to change it to.

  • Character sheets

    Use a character reference to create new poses and views.

  • 1K and 2K output

    Choose 1K and 2K in 7 aspect ratios: 1:1, 4:3, 3:4, 3:2, 2:3, 16:9 and 9:16.

  • Reference inputs

    Besides a text prompt you can add up to 10 reference images.

  • Up to 4 images per run

    Generate 1–4 variations at once and keep the best; each image is billed and refunded on its own.

Use cases

Starting points you can copy into the generator above.

Use caseGood forExample prompt
Posters with textReadable headlines in Chinese or English.

A minimalist poster for a tea festival with the title "Autumn Tea Fair" in elegant brush lettering, ink wash mountains, cream paper texture

Portrait editsChange outfits or backgrounds.

Change the jacket in the photo to a navy wool coat and keep the face, pose and lighting the same

Multi-image compositesCombine elements from several images.

Put the person from image 1 in the cafe from image 2, holding the cup from image 3, natural lighting

Character turnaroundsFront, side and back views.

Based on the character reference, draw a three-view sheet: front, side and back, plain white background

Qwen Image 2.1 in 3 steps

  1. 1

    Write a prompt or add references

    Describe subject, style, light and any text in quotes. To edit, upload reference images and say exactly what should change.

  2. 2

    Pick size, aspect ratio and count

    Choose the resolution, aspect ratio and 1–4 images. The credit cost is shown before you submit.

  3. 3

    Generate and keep your results

    Images appear in the generator and are saved to My works. Anything that fails is refunded automatically.

X Posts About Qwen Image 2.1 — Open Image Generation and Editing in One Model

YouTube Videos About Qwen Image 2.1 — Open Image Generation and Editing in One Model

Frequently asked questions

Yes. Alibaba released it as an open 7B model; here you can use it without any setup.

Ready to Edit with Qwen Image 2.1?

Text-savvy generation plus reference editing with up to 10 images, at 2K.

Explore All Models
  • Up to 2K output
  • Up to 10 reference images
  • Failed outputs refunded