Create structured infographics, multilingual campaign visuals, reference-guided compositions, and focused image edits with a controllable 1K or 2K workflow.
Built for visual briefs that need structure, language, references, and controlled revision.
Structured information
Organize components, steps, and relationships into clear diagrams and visual explainers.
Multilingual text
Direct short campaign copy in several languages while keeping one coherent visual hierarchy.
Reference fusion
Combine products, materials, packaging, and scene references into one coordinated result.
Focused editing
Describe a targeted change while preserving the product, framing, and surrounding details.
Key Features
Four Seedream strengths for controlled visual creation.
Each capability maps to a workflow Nanobana currently supports, from starting with text to refining an image with several references.
Feature 01 · Information visualization
Turn dense instructions into a visual people can follow.
Seedream 5.0 Pro is designed to organize data, concepts, and structured relationships into clear visual layouts, making it useful for diagrams, guides, and explanatory graphics.
Arrange steps and components in a visible order
Give labels and illustrations a clear hierarchy
Translate a detailed brief into one coherent explainer
Feature 02 · Native multilingual text
Build localized visuals inside one consistent design.
The model supports native text generation in 14 languages, helping creators explore multilingual posters, packaging, and campaign concepts. Generated copy still needs careful proofreading.
Combine concise phrases from several languages
Direct placement, emphasis, and visual hierarchy
Keep the same campaign language across localized ideas
Feature 03 · Multi-reference fusion
Bring separate visual ingredients into one scene.
Seedream 5.0 Pro can use several reference images to combine products, materials, people, and environments. Nanobana's quick generator accepts six references, while the full Generator supports up to ten for this model.
Assign a clear role to each uploaded reference
Carry product shapes, colors, and materials forward
Compose a new result instead of simply collaging inputs
Feature 04 · Instruction-led editing
Revise one part without rebuilding the whole image.
Use a reference image and a precise natural-language instruction to change an object, color, background, or lighting condition while asking the model to preserve everything that already works.
State exactly what may change
Repeat the product and composition details to preserve
Refine one focused variable at a time
Examples + prompt recipes
Explore four Seedream workflows through one product campaign.
These fictional KADO lamp concepts demonstrate information design, multilingual typography, reference-guided composition, and a controlled image edit. Use the prompts as starting points and inspect every generated detail.
A four-step assembly guide
A modular product becomes an ordered visual explanation with concise, inspectable labels.
View the prompt+
Design a clean 3:2 illustrated assembly guide for a fictional modular desk lamp called "KADO". Use a warm ivory background and show four numbered stages from left to right with simple exploded views. Render only the exact labels "1 · BASE", "2 · ARM", "3 · SHADE", and "4 · LIGHT". The lamp has a brushed aluminum body, amber glass shade, and one cobalt control ring. Use precise arrows, generous spacing, dark charcoal linework, and no additional text.
Short exact phrases share a restrained campaign hierarchy without competing with the product.
View the prompt+
Create a refined 3:2 launch poster for a fictional modular desk lamp named "KADO". Show the brushed aluminum lamp with an amber glass shade on a warm ivory architectural surface. Render only the exact text "KADO", "LIGHT, YOUR WAY.", "自在光明", and "光を整える". Use a quiet Swiss editorial grid, one cobalt-blue accent, soft directional light, generous whitespace, and no additional text.
Product, material, packaging, and interior cues carry into a coordinated family of applications.
View the prompt+
Using the uploaded KADO lamp, amber glass, brushed aluminum, packaging, and interior references, create one polished 3:2 campaign presentation. Include the same lamp, a compact floor-lamp variation, warm ivory packaging, a small catalog spread, and one quiet architectural room scene. Preserve the amber shade, cobalt control ring, aluminum finish, product proportions, and KADO identity across every application. Use soft studio light and do not add unrelated products or brands.
A focused edit moves a daylight product scene to blue hour without redesigning the lamp.
View the prompt+
Using the uploaded KADO lamp photograph, change only the environment from soft morning daylight to a quiet blue-hour library with one warm pool of light from the lamp. Keep the lamp body, amber glass shade, cobalt control ring, camera angle, crop, scale, furniture geometry, and shadow footprint unchanged. Preserve realistic aluminum and glass materials. Do not add text or new objects.
Move from a quick brief to a downloadable Seedream image.
Begin on the model page, then continue in Nanobana's established Generator with your model, prompt, mode, and references carried forward.
01
Write or add references
Describe a new image, or add up to six references in the quick generator when a product, material, or composition must carry forward.
02
Continue with Seedream 5.0 Pro
Choose Quick Generate to open the full image workflow with Seedream 5.0 Pro, your prompt, mode, and references preserved.
03
Set, generate, and review
Choose a supported ratio, 1K or 2K quality, and PNG or JPEG output, then generate, inspect, refine, and download the strongest result.
Seedream 5.0 Pro questions
What is Seedream 5.0 Pro?+
Seedream 5.0 Pro is a multimodal image generation and editing model developed by ByteDance. It accepts text and optional image references and produces a single image result.
What is the relationship between Nanobana, Seedream 5.0 Pro, and ByteDance?+
Seedream 5.0 Pro is developed by ByteDance. Nanobana is an independent creative platform that provides a web workflow for supported third-party models, including Seedream 5.0 Pro. Nanobana does not develop or own the model and is not affiliated with or endorsed by ByteDance.
Do I need a ByteDance or Dreamina account to use it on Nanobana?+
No separate ByteDance or Dreamina login is required for this workflow. Sign in to Nanobana and use the model through the Nanobana Generator, subject to the credits and plan shown in your Nanobana account.
Which Seedream 5.0 Pro modes does Nanobana support?+
Nanobana supports text-to-image and image-to-image workflows. Start with only a prompt for a new image, or upload one or more references when the result should preserve or combine specific visual material.
How many reference images can I add?+
The model-page quick generator accepts up to six references. After continuing to the full Generator, Seedream 5.0 Pro supports the current provider-backed limit of ten reference images. Use clear, complementary inputs and explain the role of each one.
Which output settings are available?+
Nanobana currently exposes 1K and 2K quality, PNG and JPEG output, and eight aspect ratios for Seedream 5.0 Pro: 1:1, 4:3, 3:4, 16:9, 9:16, 2:3, 3:2, and 21:9.
Can Seedream 5.0 Pro generate readable multilingual text?+
The model is designed for native text generation in 14 languages, but spelling, punctuation, translation, and small labels can still be wrong. Keep required copy short, place exact phrases in quotation marks, and proofread every result before publishing.
How should I prompt an infographic or focused edit?+
For an infographic, define the audience, layout order, exact labels, and relationships, then verify every fact and number. For an edit, state what may change and repeat the subject, framing, materials, and surrounding details that must remain unchanged.
Start a controlled Seedream 5.0 Pro image.
Write your own brief or begin with one of the KADO prompts above. The generator is ready with Seedream 5.0 Pro selected.
