Qwen Image Family
Generated on PixelDojo. Produced by PixelDojo's generation pipeline.
Every Qwen image model here costs either 1 or 2 credits a picture, and editing starts at 1. Three generations are live: the original Qwen Image line with its Plus and Max tiers, the Qwen Image 2.0 line with a standard and a Pro tier, and Qwen Image 3.0 Pro, which generates and edits from one endpoint and prices by output size instead of by tier. Below is each entry with its credit price, how many images it returns per call, and how many reference images it accepts. Every figure was read from the model configuration files on August 26, 2026.
The Qwen image line, entry by entry
Qwen Image Plus
1 credit per image. The fast lane of the original Qwen Image line and the default when no model is named. It is one of the few entries in the whole catalog that also loads a custom LoRA, with its own strength, step count and guidance controls.
Qwen Image Max
2 credits per image. The higher quality tier of the same line, at double the price and with the same aspect ratio options: square, 16:9, 9:16, 4:3, 3:4 and 21:9.
Qwen Image 2.0
1 credit per image, and a single call returns up to 6 images at 1 credit each. Known for sharp in image text and style transfer. Sizes are free form as long as the total pixel count sits between 512 squared and 2048 squared.
Qwen Image 2.0 Pro
2 credits per image, up to 6 per call. Same size freedom as the standard 2.0 tier, with the extra fidelity that the Pro build brings. This is the tier to reach for when a sign or a label in the frame has to be readable.
Qwen 3 Pro
Qwen Image 3.0 Pro prices by output size rather than by tier: 1 credit for a 1K image and 2 for a 2K one. Leaving the size unset lets the model choose and bills at the 2 credit rate. It generates from a prompt alone and edits from 1 to 3 supplied images on the same endpoint, up to 6 outputs per call.
Qwen Image Edit Plus
1 credit per output image. Supply 1 to 3 reference images with an edit instruction and it returns up to 6 results in one call, priced per image. Omit the size and the output matches the input's aspect ratio.
Qwen Image Edit Max
1.5 credits per output image, the only price in the Qwen family that is not a whole number. Same 1 to 3 reference images and same 1 to 6 outputs per call as Edit Plus.
Qwen Image Edit, the original
1 credit, and it is fixed at a single output image per call with no size control. Kept for callers pinned to it. Both newer edit tiers do more for the same or nearly the same price.
Qwen Image 2 Edit
The edit sibling of the 2.0 line, priced at its parent's rate: 1 credit per image on the standard tier and 2 on Pro. It takes 1 to 3 reference images, and when no size is given the output aspect ratio follows the last image in the list, so ordering matters.
Sizes and aspect ratios
The 2.0 line accepts any dimensions with a total pixel count between 512 squared and 2048 squared, exposed as 1K, 2K and a maximum tier close to 2048 by 2048. Qwen Image 3.0 Pro offers two tiers, roughly 1 and 2 megapixels, across eight ratios: 1:1, 2:3, 3:2, 3:4, 4:3, 9:16, 16:9 and 21:9.
Rolling names and dated pins
Each family publishes a rolling name plus one or two dated snapshots. The rolling name always points at current behaviour. Pin a dated id when you need a render to stay reproducible, and it costs exactly what its rolling parent costs.
Training your own Qwen model
The Qwen Image LoRA trainer is 25 credits per 1000 steps, so a 3000 step run is 75 credits. Training takes roughly 20 to 40 minutes and the finished model is yours to generate from.
Nine Qwen image models across three generations. Nothing in the family costs more than 2 credits a picture, and the cheapest edit is 1 credit.
Why Choose Pixel Dojo for Qwen Image Family
Professional-quality results with cutting-edge AI technology
The whole family fits in a 1 to 2 credit band
Plus, 2.0, Edit Plus, the legacy entries and a 1K render on 3.0 Pro are all 1 credit. Max, 2.0 Pro and a 2K render on 3.0 Pro are all 2. Edit Max at 1.5 is the only figure in between.
Six images from a single call
The 2.0 line, the edit tiers and 3.0 Pro all accept a count of up to 6, priced per image. That is the cheapest way to compare six compositions before committing to one.
Text in the frame is the family strength
Sharp in image text is what the 2.0 line is built for, and 3.0 Pro carries it into the newest generation. If the shot contains a sign, a label or a packaging mock up, this is the family to try first.
Editing is priced like generating
There is no premium for supplying reference images. An edit on the standard tiers costs the same 1 credit as a fresh generation, so iterating on an existing picture is not more expensive than starting over.
How It Works
How to pick a Qwen entry and know the bill in advance:
Start from generate or edit
A prompt on its own goes to Plus, Max, 2.0, 2.0 Pro or 3.0 Pro. A picture plus an instruction goes to one of the edit entries, or to 3.0 Pro, which handles both from one endpoint.
Pick the tier, which is the price
Standard tiers are 1 credit an image, Pro tiers are 2, and Edit Max is 1.5. On Qwen Image 3.0 Pro the size does the same job: 1K costs 1 credit and 2K costs 2.
Ask for several images at once
The count field goes to 6 on the newer entries and each image is charged separately. Six variations at 1 credit each is one call and 6 credits, which beats six separate runs on wall clock time.
Loved by creators on PixelDojo
Real feedback from people using PixelDojo, pulled from our in-product surveys.
Great site and so much fun to use!
very useful set of tools for image creation, upscaling and enhancement
ease of use, variety of tools, high quality trainings, and a well-maintained discord channel
Ease of use, friendliness and support of the owner, continued innovation.
it is an amazing site to create a pics and vids for those who don't have the hardware themselves
good tools in one place
Explore more AI tools on PixelDojo
AI Tools
Compare & Switch
- Best AI Image Generators
- Best AI Video Generators
- Midjourney Alternatives
- Civitai Alternatives
- Runway Alternatives
- Leonardo Alternatives
- Pika Alternatives
- Luma Alternatives
- Magnific Alternatives
- Veo Alternatives
- Flux Alternatives
- Freepik Alternatives
- Seedance Alternatives
- Seedream Alternatives
- Pixverse Alternatives
- GPT Image Alternatives
- Synthesia Alternatives
- Playground Alternatives
- NightCafe Alternatives
- Canva AI Alternatives
- ElevenLabs Alternatives
- ComfyUI Alternatives
- Fal Alternatives
- Replicate Alternatives
Common Questions
Everything you need to know about Qwen Image Family
How much does a Qwen image cost?
1 or 2 credits per image. Qwen Image Plus, Qwen Image 2.0 and a 1K render on Qwen Image 3.0 Pro are 1 credit. Qwen Image Max, Qwen Image 2.0 Pro and a 2K render on 3.0 Pro are 2. Editing is 1 credit on the standard tiers and 1.5 on Edit Max.
What is the difference between Qwen Image Plus and Max?
Price and quality tier. Plus is 1 credit an image and Max is 2, with the same six aspect ratios on both. Plus is also the entry that accepts a custom LoRA, so style work with your own trained model belongs there.
Which Qwen model should I use for text in an image?
The 2.0 line is built for sharp in image text, and Qwen Image 2.0 Pro at 2 credits is the strongest of that pair. Qwen Image 3.0 Pro carries the same strength into the newest generation and costs 1 credit at 1K.
Can Qwen models edit an existing image?
Yes, several. Qwen Image Edit Plus is 1 credit, Edit Max is 1.5, Qwen Image 2 Edit charges its parent tier's rate, and Qwen Image 3.0 Pro edits from the same endpoint it generates on. All of them take 1 to 3 reference images with an instruction.
What sizes do Qwen image models output?
The 2.0 line takes any dimensions whose total pixel count sits between 512 squared and 2048 squared, offered as 1K, 2K and a maximum tier near 2048 by 2048. Qwen Image 3.0 Pro offers a 1K and a 2K tier across eight aspect ratios, and the tier you pick is what sets the price.
Can I train a custom Qwen model?
Yes. The Qwen Image LoRA trainer charges 25 credits per 1000 steps, so a 2000 step run is 50 credits and a 3000 step run is 75. Most runs finish in 20 to 40 minutes.
How do I call a Qwen model from the API?
POST to /api/v1/models/{apiId}/run with your key, using ids such as qwen-image-plus, qwen-image-max, qwen-image-2, qwen-image-2-edit or qwen-3-pro. Same key and same request shape as every other model in the catalog.