Qwen
Generate and edit images on Qwen with sharp text, natural retouching, and seamless object swaps. Describe the change to invideo’s agent and it writes the instruction the model executes.
What sets Qwen apart
Edits that follow instructions
Qwen edits from a sentence: add, remove, or move an object, change a pose, swap a background, restyle the scene. You describe the change and the model performs it, no masks, no layers, no manual tools.
Great facial consistency
Identity holds through the edit. Change the outfit, the lighting, or the background, and the person still looks exactly like themselves, in solo portraits and in group photos where every face has to survive.
New angles from one photo
Qwen generates the same subject from different camera perspectives: rotate to a side view, go top-down, pull wider. Identity, materials, and lighting stay coherent, so the new angle reads as the same object, not a redraw.
Add up to three images as references
Feed multiple images into a single instruction: the person from one, the outfit from another, the pose from a third. The model composes across them instead of editing them one at a time.
It edits the text inside your image
Text inside an image can be rewritten while matching the original font, size, and style, in English and Chinese. The sign says what you need it to say, and it still looks like the sign.
What Qwen is best used for
Retouching
Qwen smooths skin, fixes blemishes, and cleans up portraits while keeping natural texture intact, so the person looks like themselves on a good day, not a render.
Product shots
Drop a product into a scene with correct lighting and depth, swap the background, or generate the missing angle. A full product gallery from one hero photo, no reshoot in sight.
Fixing the photo you already have
Harsh light spots, unwanted objects, a background that fights the subject: described in a sentence, gone in a pass. The photo you almost got becomes the photo you needed.
How to use Qwen with invideo agents
On invideo, Qwen runs inside an agentic workflow: the agent takes your note, picks the right Qwen tool for it, and writes the instruction the model executes. Here is how it works in practice:
The agent picks the right model intelligently.
Qwen sits on a roster of 200+ models, and the agent routes work to it where it wins: precision edits, retouching, product placement, angle changes, and any fix that should not cost a regeneration.
Your note picks the tool, so you never have to.
Qwen on invideo is a set of specialized edit tools, and the agent knows which one your note calls for: say the lighting is harsh and it runs the relight, say you need the side view and it runs multi-angle. You describe the problem; the tooling is not your problem.
The fix touches only what you flagged.
The agent writes Qwen's instruction to the exact change you asked for, and the rest of the image comes back untouched. The photo you liked is still the photo you liked, minus the thing that bothered you.
You always stay in control.
You set how much the agent does on its own: let it generate images freely, or see every prompt before it hits generate. That setting is yours to make, and yours to change.
Helping creatives stay creative
Multiplayer mode
Collaborate in real time with live cursors to show what everyone's working on.
Storyboarding
Turn any script or idea into a shot-by-shot plan, then tweak as needed before generating.
Script writing
Write your script inside invideo, and ask an AI co-writer for help if you'd like.
Timeline editor
Picture Premiere Pro with full AI.
Build your own agents
Create custom agents to fill specific roles like cinematographer, music designer, and more.
From solo creatives to creative enterprises
World-class investors stand behind invideo.
Backed by the firms behind Stripe, Spotify, Flipkart, and ByteDance.
Pricing
Access to 200+ image, video, audio, music models including Seedance 2.0, Veo 3.1, Kling 3.0, Nano banana pro & Elevenlabs music.
Access to top stock providers like iStock, Storyblocks & more.
Model & agent prices are subject to change.
On-demand credit top-ups available.
Qwen FAQs
What is Qwen?
Qwen is Alibaba's family of AI models. On invideo, Qwen refers to its vision models for images: generation with exceptional text rendering, and precision editing that changes exactly what you ask, from lighting and skin to objects, backgrounds, and camera angles.
What is Qwen Edit used for?
Instruction-based photo editing: retouching skin, refining portraits, adjusting people in group photos, adding products, removing objects and light spots, relighting scenes, swapping backgrounds, and generating new angles of the same subject.
How is Qwen Edit different from Qwen Image?
Qwen Image generates new images from text; Qwen Edit changes images you already have. On invideo, both sit behind the agent, and your request decides which one runs.
What types of edits can Qwen Edit perform?
Two kinds: appearance edits that change specific elements while leaving the rest untouched (objects, colors, backgrounds, light), and semantic edits that transform the image while keeping its identity (style transfer, pose changes, new angles).
Do I need editing skills to use Qwen Edit?
No. There are no masks, layers, or tools to learn: you describe the change in plain language, and on invideo the agent writes the instruction and picks the right Qwen tool for it.
Can Qwen Edit work with more than one image?
Yes, up to three inputs in one instruction: take the person from one image, the outfit from a second, and the pose from a third, and compose them into a single result.
Does Qwen support text changes inside images?
Yes. It rewrites text inside an image while matching the original font, size, and style, in English and Chinese, so signs, labels, and posters stay believable after the edit.
Where can I access Qwen Edit tools?
On invideo, under Agents & Models, or by just telling the agent what to fix; it selects the right Qwen tool for the job. Qwen models are also available through Alibaba's own platforms.