Key Capabilities
Discover the powerful features of Seedream 4.5
Advanced Text Understanding
Seedream 4.5 has significantly stronger text comprehension compared to previous versions. Generate high-quality images with concise prompts - simple, clear language works better than complex keyword stacking.
Versatile Image Editing
Perform precise image editing operations including addition, deletion, replacement, and modification through text prompts. Use visual indicators like arrows and bounding boxes to specify editing regions when needed.
Multi-Image Creation
Support for multi-image input and output with consistent character continuity and unified style. Perfect for storyboarding, comic creation, and design scenarios requiring cohesive visual identity.
Master Your Prompts
Learn best practices and guidelines for creating effective prompts
Clearly Describe the Scene Using Natural Language
Use coherent natural language to describe the subject, action, and environment. If aesthetics matter, include descriptors of style, color, lighting, or composition.
Subject
The main character or object in your image
Action
What is happening or the state of the subject
Environment
The setting, background, and surrounding context
Aesthetics (Optional)
Style, color palette, lighting conditions, composition
💡 Tips:
- • Use coherent natural language instead of comma-separated keywords
- • Seedream 4.5 has strong text understanding - concise prompts work better than complex stacking
- • Avoid over-describing; the model understands natural sentences
- • Include style descriptors only when specific aesthetics are important
✅ Recommended
A girl in a lavish dress walking under a parasol along a tree-lined path, in the style of a Monet oil painting.
❌ Not Recommended
Girl, umbrella, tree-lined street, oil painting texture.
Specify Application Scenario
If you have a specific use case, explicitly state the image purpose and type in your prompt. This helps Seedream 4.5 understand the context and generate more appropriate results.
Purpose
What will the image be used for (logo, poster, icon, illustration, UI mockup)
Type
The specific category or format of the output
Requirements
Any technical or design requirements specific to the use case
💡 Tips:
- • Explicitly state the intended use (e.g., 'Design a logo for...')
- • Specify output type (logo, poster, UI mockup, infographic, storyboard)
- • Include technical requirements if needed (layout, composition, format)
- • Mention the target audience or industry when relevant
✅ Recommended
Design a logo for a gaming company. The logo features a dog playing with a game controller. The company name 'PITBULL' is written on it.
❌ Not Recommended
An abstract image of a dog holding a controller, and the word PITBULL on it.
Improve Text Rendering Accuracy
Use double quotation marks for the text that needs to appear in the generated image. This significantly improves text rendering accuracy in Seedream 4.5.
Quotation Marks
Wrap desired text in double quotes to improve accuracy
Context
Specify where and how the text should appear
Style (Optional)
Describe the visual style of the text if important
💡 Tips:
- • Always use double quotation marks around text that must appear exactly
- • Specify text placement (on poster, on sign, on product, on screen)
- • Be clear about text styling if it matters (font style, size, color)
- • Keep text concise for better rendering accuracy
✅ Recommended
Generate a poster with the title "Seedream 4.5".
❌ Not Recommended
Generate a poster titled Seedream 4.5.
Clearly Define Image Editing Goals and Fixed Elements
Use concise, unambiguous instructions for modifications. Avoid vague pronouns. If other elements should remain unchanged, specify that explicitly.
Target Element
Precisely identify what to edit (avoid vague pronouns like 'that one' or 'it')
Modification
Clearly describe the desired change (add, remove, replace, modify)
Fixed Elements
Specify what should remain unchanged
💡 Tips:
- • Use specific descriptors, not vague pronouns ('the tallest panda' not 'that one')
- • Explicitly state what should remain unchanged (pose, expression, background)
- • Use visual indicators (arrows, bounding boxes, doodles) for complex edits
- • Be concise but unambiguous in your instructions
✅ Recommended
Dress the tallest panda in pink Peking Opera costume and headgear, keeping its pose unchanged.
❌ Not Recommended
Put that one in pink clothes.
Prompt Examples & Image Results
🖼️ Text-to-Image
| Type | Prompt Examples | Image Result |
|---|---|---|
| Complex Scene Description | A cluttered office desk. On the desk, there is an open laptop with a screen displaying green code. Next to it, a mug with the word 'Developer' on it, with steam rising from the top. An open book lies on the desk, with pages showing a Venn diagram illustrating the nesting relationships of three circles in gray, blue, and light green. A sticky note with a mind map drawn on it, organized in a three-level vertical structure. A fountain pen, with the cap lying beside it. Next to the pen is a smartphone, with a new message notification displayed on the screen. In the corner of the desk, there is a small pot of succulent plants. The background is a blurred bookshelf. Sunlight shines from the right side, casting light and shadow on the desk. | |
| Detailed Layout Control | Interior view of an open refrigerator: Top shelf: On the left, there is a carton of milk featuring an illustration of three cows of different sizes grazing on a grassland. On the right, there is an egg holder containing eight eggs. Middle shelf: A plate holds leftover roasted chicken with a small red flag stuck into it. Next to it is a transparent container filled with strawberries, the container is decorated with images of a pineapple, strawberries and oranges. Bottom shelf: The vegetable drawer contains lettuce, carrots and tomatoes. On the door shelves, there are bottles of ketchup and mayonnaise. | |
| Knowledge Visualization | Draw the following system of binary linear equations and the corresponding solution steps on the blackboard: 5x + 2y = 26; 2x - y = 5. |