How does interactive editing work in Seedream 5.0 Pro?

You add a marker on the reference image - a coordinate point, a freehand sketch, or a lassoed area - and pair it with a text prompt. The model reads the marked location and edits only that area, blending the change naturally into the original scene.

Draw on your image. Describe the edit. Done.

Q: Can I combine several images into one?

Yes. Multi-image fusion accepts up to 10 reference images plus a text instruction, and merges their styles or elements into a single new image.

Q: Is the editor free to use?

You can start free on Fullmira AI with a daily credit allowance and no credit card. Heavier use is covered by paid plans you can upgrade to or cancel anytime.

Drop a point, sketch, or lasso right on the picture, write what you want there, and Seedream 5.0 Pro edits only that spot - blending it naturally into the scene. No signup to try.

Seedream 5.0 Pro

Lasso, sketch and point editing

Fuse up to 10 references

1K / 2K - PNG or JPEG

How it works

Precise editing in three moves

Seedream 5.0 Pro reads marks you place on the image and edits exactly there - no re-rolling the whole picture.

Bring your image

Upload a photo or pick an image you have already made. The model works from your reference directly.

Mark the spot

Drop a coordinate point, doodle a rough sketch, or lasso an area. The mark tells the model precisely where to work.

Describe and generate

Write the change in plain words. The model edits inside your mark and blends it into the original scene, then outputs at 1K or 2K.

Features

What Seedream 5.0 Pro does best

Built for precise, controllable production edits from your own reference images.

Interactive editing - marks and coordinates

Mark it, then say what changes.

Overlay a coordinate point, a freehand sketch, or a lasso on your reference image and pair it with a prompt. The model maps the mark to an exact location and performs a local edit while the rest of the frame is preserved.

Coordinate points for pinpoint object placement
Freehand sketch and lasso to define any area
New content blends naturally into the scene

Multi-image fusion

Blend up to 10 references into one image.

Feed the model several reference images plus a text instruction, and it fuses their styles and elements into a single new image - swap an outfit onto a model, place a person into a scene, or combine product shots.

Up to 10 reference images per generation
Material, style and element transfer between photos
Text-to-image and image-to-image in one tool

Made with Fullmira

What people create and edit

Real images generated by Seedream 5.0 Pro — portraits, fashion edits, surreal composites.

iPhone 16 Pro ski ad, mirrored goggles reflecting a girl in pink jacket taking a selfie, snowy mountains backdrop

3D social media profile mockup, girl with round glasses bursting through a torn screen, 4K AI art style

Retro Polaroid photo, couple embracing against white curtains, cool low-saturation tones, film grain, pink border

Wide-angle tracking shot, young skater on graffiti street at golden hour, 35mm film look, high contrast

Young man in white shirt and bright pink blazer holding a black lamb, golden side light, Kodak Portra 400

Surrealist fashion shoot, student surrounded by floating giant clay heads and face-books, pastel macaron tones

Candid portrait, 21-year-old blonde in pink velvet tracksuit with butterfly clips, Y2K aesthetic, golden hour

Ultra-wide low-angle portrait, young woman leaning toward camera, noon sunlight with lens flare, blue sky, 35mm fisheye style

Pro vs Lite

Which Seedream fits your job?

Pro is the high-control editor; Lite is the fast, high-resolution generator.

Capability	5.0 Pro	5.0 Lite
Mark-based interactive editing (point / sketch)	Yes	-
Multi-image fusion	Up to 10 refs	Up to 14 refs
Text-to-image and image-to-image	Yes	Yes
Grouped image output	Single image	Yes
Resolution	1K / 2K	2K / 3K / 4K
Output format	PNG / JPEG	PNG / JPEG

FAQ

Questions, answered

How does interactive editing work?

You add a marker on your reference image - a coordinate point, a freehand sketch, or a lassoed area - and pair it with a text prompt. Seedream 5.0 Pro reads the marked location and edits only that area, blending the new content naturally into the original scene.

Can I combine several images into one?

Yes. Multi-image fusion takes up to 10 reference images plus a text instruction and merges their styles or elements into one new image, such as putting the outfit from image 2 onto the model in image 1.

What output does it support?

Seedream 5.0 Pro outputs a single image at 1K or 2K resolution, in PNG or JPEG, with an optional watermark. It focuses on high-quality single-image editing rather than grouped output or streaming.

Any tips for writing the prompt?

Write clear, natural language covering subject, action, and setting. For edits, state what to add or change and ask it to keep the composition and remove any sketch lines.

Is it free to use on Fullmira AI?

You can start free with a daily credit allowance and no credit card. Heavier use is covered by paid plans you can upgrade to or cancel anytime.

Free to start - No credit card

Mark it. Describe it. Ship it.

Precise, mark-based image editing and multi-image fusion - all in one place on Fullmira AI.

Mark-based edits10-image fusionSwitch models anytime