Skip to main content
Kyeo AI
AI image model
ByteDance
Seedream

Seedream 4.5

Seedream 4.5 / Capabilities — ByteDance's official Seedream 4.5 material highlights multi-image editing, preservation of reference details, layout, and text rendering. The Kyeo workflow has its own concrete limits: prompts can contain up to 3,000 characters, 8 aspect ratios are available, and editing accepts up to 14 JPG, PNG, or WebP images at 10 MB each. Kyeo AI currently charges a fixed 7 credits per request, with the same price for basic and high. Budget for retries, revisions, and the share of outputs that pass review.

Use this modelView credit pricingBack to Model Hub
Capabilities
Text-to-image, image-to-image
Credit cost
7 credits per request
Prompt limit
3,000 characters
Upload limit
Up to 14 images for edits
30-second overview
  • Seedream 4.5 / Capabilities — ByteDance highlights multi-image editing, reference-detail preservation, layout, and text rendering.
  • Seedream 4.5 / Integration — Kyeo supports text-to-image and reference editing through one model entry.
  • Seedream 4.5 / Credit cost — Kyeo AI currently charges a fixed 7 credits per request, with the same price for basic and high. Budget for retries, revisions, and the share of outputs that pass review.
  • Seedream 4.5 / Aspect ratio / Quality — Choose from 8 aspect ratios and basic 2K or high 4K output.

Popular searches: using Seedream 4.5 on Kyeo AI · how much a Seedream 4.5 request costs · which Seedream 4.5 controls are available

Examples

Visual concept

Visual concept

FORMAT: 4:5 vertical premium smartphone campaign poster, ultra-high resolution (8K), global OOH + digital + social ready Style: Apple-level minimalism × bold Gen-Z composition × hyper-real commercial photography 🧠 CORE INTENT: “PRO, WITHOUT TRYING.” Effortless power. Quiet confidence. No noise — just precision. 🎬 SCENE COMPOSITION: BACKGROUND: Pure matte white canvas with a dominant rounded rectangle block in ultra-vibrant burnt orange gradient (top-left → bottom-right glow) GIANT TYPOGRAPHY (BACKGROUND LAYER): “PRO” - ultra-bold geometric sans-serif - stretched wide kerning - partially cropped by frame edges - soft shadow depth - color: deep orange slightly darker than background block - subtle emboss effect 👤 SUBJECT (HUMAN ELEMENT): - Female model, clean editorial look (Apple casting style) - Neutral expression (confident, calm, not smiling) - Hair: natural flow, slightly wind-touched - Outfit: → muted tones (off-white + deep maroon jacket) → minimal styling, no distractions POSE: - Shot from slightly low angle (hero perspective) - One hand extended toward camera holding iPhone - Body slightly angled for depth - Face aligned toward viewer (eye contact) 📱 PRODUCT (HERO FOCUS): DEVICE: iPhone 17 Pro in Titanium Orange finish POSITION: - Extreme foreground (forced perspective) - Slight tilt to reveal camera module depth DETAILING: - triple camera system hyper-sharp - lens reflections physically accurate - brushed titanium edge highlights - micro-scratches + realistic material response - Apple logo centered, subtle reflective gloss REFLECTION: - soft studio reflection on device edges - micro light bloom around camera rings ✨ LIGHTING SYSTEM: MAIN LIGHT: Soft studio key light (front-left) → smooth skin tones → clean product highlights RIM LIGHT: Warm edge light from right → enhances phone silhouette → highlights metallic edges ACCENT LIGHT: Subtle glow from background orange block SHADOW: Soft drop shadow under subject + phone (no harsh contrast) 🎨 COLOR SYSTEM: - Primary: Apple Orange (hero) - Secondary: Clean white - Accent: warm highlights + skin tones - Contrast: deep micro-shadows No over-saturation. No neon. Controlled premium palette. ✨ GRAPHIC ELEMENTS: - Minimal sparkle icons ✦ (2–3 only) - Soft floating UI pill (glass morphism) - Very subtle grain for realism ✍️ TYPOGRAPHY SYSTEM: TOP LEFT: Apple logo (black, small, clean) TOP RIGHT: “Built for Apple Intelligence.” (thin sans-serif) MID LEFT (pill): “iPhone 17 Pro” CENTER DOMINANT: “PRO” (background type) BOTTOM RIGHT (body text): Small paragraph: “iPhone 17 Pro. Uncompromising power. A19 Pro chip. Advanced camera system. Designed for what’s next.” 📊 FEATURE STRIP (BOTTOM GRID): 4 rounded cards (glass-white panels): 1. A19 PRO CHIP “Unstoppable performance.” 2. 48MP PRO CAMERA “Total creative control.” 3. 4K DOLBY VISION “Cinematic in every frame.” 4. APPLE INTELLIGENCE “Personal. Private. Powerful.” Icons minimal line style 📐 COMPOSITION GRID: - Top: brand + statement - Mid: subject + product (dominant) - Background: typography layer - Bottom: feature modules Eye flow: Face → Phone → Camera → “PRO” → Features 🎥 CAMERA SPECS: - ARRI Alexa Mini LF - 50mm anamorphic lens - shallow depth of field - HDR rendering - ultra-real skin + material detail - slight lens distortion for realism

Visual concept

Visual concept

Create a 9:16 fashion editorial poster built around a real dressing-room moment. Scene: [Scene] Fashion action: [Fashion Action] Outfit: [Outfit] Accent color: [Accent Color] Headline: [Headline] Use realistic black-and-white documentary photography as the main visual. The model should be naturally interacting with the clothing — adjusting, changing, checking the fit, walking, or preparing to step out — not posing directly for the camera. Combine the monochrome photo with one single highly saturated accent color, warm off-white paper background, strong negative space, and oversized condensed sans-serif typography. Let the typography interact with the subject through cropping and layering, but keep all headline letters clean, solid, sharp, and fully rendered — no distressed, dirty, scratched, or broken text. Add only a few meaningful editorial details such as time, fitting-room status, look name, or a short phrase. Keep the layout bold, asymmetric, minimal, and fashion-magazine driven. Style: Swiss editorial design × brutalist typography × fashion documentary photography × modern print poster. Avoid multiple accent colors, random decorative text, excessive graphic symbols, messy typography, glossy commercial advertising, stiff model poses, CGI skin, and cluttered layouts.

Visual concept

Visual concept

Vintage filmstrip collage of a woman Ultra-realistic IMAX-level Netflix-style asymmetrical film-style memory collage portrait, 9:16 vertical composition, use the uploaded reference image as the primary facial reference for the young woman with maximum face consistency and highest identity preservation, maintaining the exact facial identity, facial structure, eye shape, nose, lips, hairstyle influence, skin tone and overall appearance from the reference image, no facial changes, no age changes and no body distortion, a cool style fresh and realistic portrait photography collage designed like a rolling timeline of memories, composed of overlapping film frames, contact-sheet strips and transparent negatives arranged in an artistic asymmetrical layout, featuring the same young woman with luminous fair milky skin and ultra-smooth natural skin texture, her face showing a range of soft authentic emotions including joyful laughter while walking, a thoughtful and dreamy expression while looking out the window, quiet concentration while fixing her hair in the mirror, a calm cozy mood while holding a coffee, and a gentle natural smile in a candid close-up, her hairstyle preserved from the reference image with clean detailed strands and realistic texture across every frame, each panel feeling like a different personal memory captured across one beautiful visual story, clean and vibrant styling with youthful sweetness and energetic charm, soft and bright natural front lighting creating a diffused glow across her face and making the skin appear luminous and refined, ultra-rich color with creamy skin tones, fresh pastel highlights, soft daylight whites, clean neutrals and subtle warm accents creating a premium Korean editorial aesthetic, crisp facial features and sharp hair detail in every frame, realistic film-border details, subtle grain texture, slightly overlapping negative strips and scrapbook-like layering without looking messy, shallow depth of field, soft airy atmosphere, realistic reflections, storytelling composition, premium fashion-editorial mood, captured in the style of a full-frame camera with a 50mm prime lens, f/2.0, shutter speed 1/160s, ISO 100, single-point focus, photorealistic rendering, HDR, global illumination, super clear high-resolution finish, masterpiece quality, award-winning cinematography, 8K production detail. Negative Prompt: altered identity, different face, face swap errors, bad collage layout, blurry face, low-detail hair, bad anatomy, distorted hands, extra fingers, messy film frames, unreadable frame details, cartoon, anime, CGI look, low quality, text, watermark, logo.

Visual concept

Visual concept

FORMAT: 4:5 vertical hyper-commercial JBL headphone campaign poster, ultra-high-resolution 8K, global OOH + Instagram + Meta advertising, BOLD TYPOGRAPHY–FIRST DESIGN, aggressive editorial graphic design, premium product photography, contemporary JBL advertising language ⸻ 🧠 CORE IDEA FEEL THE SOUND. Make typography the dominant visual element, with JBL’s signature orange brand color controlling the entire visual environment. The poster should communicate powerful JBL audio through a physical graphic metaphor: SOUND → IMPACT → MOVEMENT The typography appears to react to the sound, creating a visual sense of bass, rhythm and energy. ⸻ 🎬 MASTER COMPOSITION Create a highly graphic, typography-dominated composition. TOP 40% — TYPOGRAPHY DOMINATION Massive stacked headline: FEEL THE SOUND. Use extremely bold condensed sans-serif typography. The words should occupy almost the entire width of the poster. Aggressive scale. Some letters cropped by the edges. Typography should feel physically pushed by sound pressure. ⸻ 🟠 BRAND-COLOR BACKGROUND Use JBL’s signature orange as the dominant full-frame background. Strong, saturated JBL orange. Build subtle tonal variations within the orange rather than introducing unrelated background colors. Use: * Deep orange shadows * Bright orange highlights * Slight darker orange graphic blocks * Black typography * White microcopy The entire poster should immediately feel like JBL even before the logo is seen. ⸻ 🎧 PRODUCT + SUBJECT Place a realistic person wearing a JBL headphone prominently in the center/lower-middle section. The headphone must remain highly recognizable and physically accurate. Subject positioned directly within the typography composition. The person should feel energetic and immersed in the music. Natural expression. No exaggerated fashion pose. The headphone is the product hero. ⸻ 💥 TYPOGRAPHIC VISUAL METAPHOR Create oversized graphic words around the product: BASS BEAT POWER RHYTHM ENERGY These words should appear as huge background typography. Some letters can stretch, overlap and partially disappear behind the subject. Use typography to create the sensation of sound physically moving through the poster. Add subtle directional typographic distortion around the headphone. No literal sound waves. No futuristic holograms. ⸻ ✍️ TYPOGRAPHY SYSTEM Typography is the hero graphic element. PRIMARY HEADLINE FEEL THE SOUND. Extremely heavy bold grotesk / condensed sans-serif. Huge scale. Tight kerning. Strong black typography against JBL orange. SECONDARY COPY JBL PURE BASS. PURE ENERGY. MICROCOPY WIRELESS • IMMERSIVE • POWERFUL Small uppercase commercial typography. ⸻ 📦 FEATURE STRIP Bottom section contains a highly structured commercial information bar: JBL PURE BASS SOUND WIRELESS FREEDOM IMMERSIVE AUDIO ALL-DAY PLAYTIME Use compact uppercase typography separated by thin black rules. Keep the information visually organized and highly legible. ⸻ 📣 CTA Large bold CTA: TURN IT UP. Below: DISCOVER JBL AUDIO JBL logo positioned prominently but cleanly. ⸻ 🎨 COLOR SYSTEM PRIMARY: JBL signature orange. SECONDARY: Deep black. ACCENT: White. Orange must dominate the entire background. Black provides the major typography contrast. White is reserved for small supporting information and selected graphic details. No unnecessary blue, purple, cyan or neon gradients. ⸻ 💡 LIGHTING High-end commercial product photography. Strong directional studio lighting. Controlled highlights on the headphones. Natural skin texture. Subtle orange environmental bounce light. Deep controlled shadows. High contrast. The product must separate clearly from the orange background. ⸻ 🔍 HYPER DETAILING Ultra-realistic headphone materials: * precise matte and gloss surfaces * realistic ear cushions * detailed controls * subtle reflections * accurate JBL branding * realistic skin pores * natural clothing texture Typography must remain razor sharp and professionally typeset. ⸻ 📐 COMPOSITION FLOW 1. MASSIVE “FEEL THE SOUND.” HEADLINE ↓ 2. JBL ORANGE GRAPHIC FIELD ↓ 3. PRODUCT + HUMAN HERO ↓ 4. GIANT BASS / BEAT / POWER TYPOGRAPHY ↓ 5. PRODUCT NAME / BRANDING ↓ 6. FEATURE STRIP ↓ 7. CTA + JBL LOGO The concept must be understood within one second. ⸻ 🎥 CAMERA / RENDER Professional commercial photography. 50mm lens. Slightly low-angle perspective to give the product presence. Ultra-realistic 8K product detail. Sharp commercial focus. Clean edges. Print-quality typography. Premium global advertising execution. ⸻ 🚫 NEGATIVE DIRECTION No generic headphone advertisement. No blue background. No purple neon. No cyberpunk aesthetic. No holograms. No floating headphones. No random particles. No excessive lens flares. No fantasy environment. No excessive cinematic VFX. No tiny headline. No minimalist typography. No luxury-fashion-only aesthetic. No cluttered unreadable layout. No incorrect JBL logo. No distorted headphones. No fake product details. ⸻ 🔥 FINAL FEEL JBL × BOLD TYPOGRAPHY × ORANGE BRAND WORLD × PRODUCT PHOTOGRAPHY × SOUND ENERGY BIG TYPE. JBL ORANGE. REAL PRODUCT. PHYSICAL SOUND. MAXIMUM COMMERCIAL IMPACT.

Visual concept

Visual concept

ADIDAS — “BREAK THE PATTERN” Format: Vertical 4:5 Social Hero Poster Style: Premium sports campaign × contemporary editorial photography × surreal architecture × bold graphic design Quality: Hyper-realistic, 8K, commercial campaign photography CORE IDEA Athletes spend years repeating the same movements until repetition becomes instinct. Instead of showing the athlete running through a normal environment, visualize routine breaking apart around them. Create a massive minimalist concrete training space where the floor is divided into hundreds of repeating geometric sections. As the athlete moves forward, those sections physically lift, rotate and rearrange into a new pathway. The world changes because the athlete refuses to follow the expected pattern. HERO ATHLETE Young professional female runner captured mid-stride. Natural athletic physique, realistic skin and subtle sweat, focused expression, loose hair reacting to movement. She wears: * Black adidas performance top * Black lightweight running shorts * White performance socks * Modern adidas running shoes with recognizable three-stripe detailing No exaggerated muscles or fashion-model posing. She should look like a real athlete captured during training. ENVIRONMENT Minimal white and concrete architectural space with enormous geometric walls and open negative space. Hundreds of rectangular floor panels create a precise grid. Some panels remain flat while others rise into ramps, walls and suspended pathways, creating a controlled impossible landscape. A thin blue accent line follows the athlete’s route through the environment. No random floating objects, fantasy scenery or psychedelic effects. COMPOSITION Keep the athlete as the visual anchor. Use the architectural grid to create strong leading lines toward her. Large portions of the poster remain clean and minimal, while selected sections contain oversized graphic elements, distance markers and cropped numbers. TYPOGRAPHY Large bold condensed typography integrated into the architecture: BREAK THE PATTERN Make PATTERN the dominant word. Supporting copy: DON’T FOLLOW THE LINE. CREATE YOUR OWN. Small technical elements: ADIDAS RUNNING / 01 PACE — PROGRESSIVE DISTANCE — 10K MODE — FORWARD Bottom CTA: RUN DIFFERENT → Add a thin blue horizontal strip: RUN / TRAIN / PUSH / REPEAT / EVOLVE PRODUCT The adidas running shoe should be highly realistic and naturally integrated into the action. Visible engineered mesh, stitching, outsole texture and realistic material deformation. The three-stripe branding should be clean and correctly proportioned without becoming oversized. LIGHTING & CAMERA Bright directional daylight entering from above, producing sharp architectural shadows and natural reflections. Professional sports photography, 50mm lens, slightly low camera position, fast shutter speed, realistic depth of field and subtle cinematic grain. Natural skin tones. Clean whites. Deep blacks. Controlled blue accent. FINAL LOOK The result should feel like a global adidas performance campaign photographed inside an impossible architectural training facility. Minimal, energetic, youthful and conceptual. No generic sneaker ad. No catalogue photography. No random AI surrealism. No excessive graphics. No luxury-fashion aesthetic. No fake CGI appearance. Vertical 4:5 · 8K UHD · hyper-realistic sports photography · surreal architecture · bold typography · premium adidas campaign · contemporary editorial art direction

Visual concept

Visual concept

Act as a senior commercial fashion art director specializing in luxury apparel product advertising and premium social media campaigns. Create an ultra-premium product advertising poster for ZARA using the EXACT reference model and the EXACT varsity jacket from the uploaded image. REFERENCE LOCK (CRITICAL): Preserve the jacket exactly as shown: — deep navy wool body — varsity yellow sleeves — embroidered chest patches exactly matching reference — yellow/navy striped ribbed collar, cuffs, and hem — same stitching, buttons, silhouette, proportions, and construction Preserve model identity exactly: — same blonde medium-length messy hair — same facial bone structure — same skin tone — same slim body proportions Do not replace or redesign the model. IMAGE INTENT: A luxury Zara global campaign poster where the varsity jacket is the undisputed hero product. The image must feel like premium commercial advertising—not editorial fashion photography. FORMAT: Vertical 4:5 Luxury product SMM poster 8K ultra-detailed OBJECTIVE: Position Zara varsity outerwear as bold contemporary street luxury with premium aspirational appeal. VISUAL HIERARCHY: Jacket = 75% importance Model = 25% importance Viewer must notice jacket before face. ART DIRECTION: Bold luxury commercial design Minimal but powerful Graphic premium composition Global brand campaign quality Behance front-page quality No catalog energy POSE: Dynamic torso twist Shoulders rotated naturally Chest angled toward camera One hand gripping collar Other hand partially inside pocket Head slightly turned away Face de-emphasized Pose exists only to showcase garment structure and silhouette. COMPOSITION: Aggressive crop from waist upward Jacket dominates frame Strong asymmetrical balance Large negative space for typography Clean premium hierarchy BACKGROUND: Ultra-premium studio backdrop Large abstract geometric planes Oversized diagonal graphic blocks Minimal visual noise Strong architectural depth COLOR SYSTEM: Primary: Deep cobalt blue #0B2D89 Secondary: Rich varsity yellow #F3B51A Support: Pure white Rich black Soft metallic silver reflections Background colors should echo jacket palette. LIGHTING: High-end commercial product lighting Sharp rim light tracing jacket outline Hard specular highlights on sleeves Controlled glossy reflections Deep sculpted shadows Premium luxury contrast FABRIC DETAIL (MOST IMPORTANT): Ultra-detailed wool grain Visible embroidery threads Crisp ribbed knit texture Natural wrinkle compression at elbows Subtle cloth tension around buttons Luxury textile realism Materials must feel tactile and expensive TYPOGRAPHY SAFE ZONE: Reserve clean headline area. Typography: ZARA Headline: MOVE DIFFERENT. Sub-copy: PREMIUM VARSITY OUTERWEAR POST PROCESS: Luxury commercial retouch Micro contrast sharpening Gloss enhancement Subtle bloom Soft halation Fine film grain Print-ready finish QUALITY: Hyper-realistic Photorealistic Luxury product advertising Premium apparel campaign Global campaign quality 8K UHD NEGATIVE: no AI slop no distorted anatomy no extra fingers no altered patches no redesigned jacket no clutter no editorial portrait no weak product focus

At a glance

Is this model right for you?

Runtime — ByteDance's official Seedream 4.5 material highlights multi-image editing, preservation of reference details, layout, and text rendering. The Kyeo workflow has its own concrete limits: prompts can contain up to 3,000 characters, 8 aspect ratios are available, and editing accepts up to 14 JPG, PNG, or WebP images at 10 MB each.

What it does best
Seedream 4.5 / Integration — ByteDance's official Seedream 4.5 material highlights multi-image editing, preservation of reference details, layout, and text rendering. The Kyeo workflow has its own concrete limits: prompts can contain up to 3,000 characters, 8 aspect ratios are available, and editing accepts up to 14 JPG, PNG, or WebP images at 10 MB each. Kyeo AI currently charges a fixed 7 credits per request, with the same price for basic and high. Budget for retries, revisions, and the share of outputs that pass review.
Who it is for
Good test cases include multi-reference composition, product or character changes, poster layouts, and briefs that continue from an existing visual direction.
Why use it on Kyeo AI
Seedream 4.5 / Capabilities — Kyeo AI keeps the practical limits visible in one workbench: 7 credits, a 3,000-character prompt, 8 aspect ratios, and up to 14 images for editing, making comparisons with the same material easier.

Key facts

Quickly evaluate whether this model fits your business use case.

Category
AI image model
Vendor
ByteDance
Model family
Seedream
Integration
Current site integration
Runtime
Asynchronous task
Prompt limit
3,000 characters
Upload limit
Up to 14 images for edits

Other names

The same model may have different names in official documentation, community discussions, and API integrations. This section brings them together for you.

Seedream 4.5
doubao-seedream-4.5

Common questions

These practical questions focus on the task, input conditions, and result requirements you should confirm before choosing.

using Seedream 4.5 on Kyeo AI
how much a Seedream 4.5 request costs
which Seedream 4.5 controls are available
setting up Seedream 4.5 text-to-image
editing with Seedream 4.5 reference images
what the Seedream 4.5 quality field means
choosing Seedream 4.5 basic or high
choosing Seedream 4.5 2K or 4K
Seedream 4.5 vs Seedream 4.0
Seedream 4.5 vs Seedream 5.0 Lite

Decision guide

Still deciding? Start with these key decision points.

1
Test it for multi-image edits

Test it for multi-image edits — Seedream 4.5 is worth sampling when a brief combines several product, character, or style references. The 14-image capacity is the interface limit, not a recommended target.

2
Compare other versions when the controls or cost differ

Compare other versions when the controls or cost differ — If explicit 1K, 2K, and 4K selection matters, compare Seedream 4.0. If you want to assess basic 2K / high 3K settings at a lower fixed Kyeo price, add Seedream 5.0 Lite.

3
Run a controlled 2K-versus-4K test

Run a controlled 2K-versus-4K test — Keep the prompt and references fixed, then run `basic` and `high`. Compare detail, text, subject preservation, and usable dimensions. Both settings cost 7 credits, and the 4K label alone does not establish content quality.

Popular comparisons

Compare common alternatives on the same task to clarify differences in inputs, controls, and cost.

What is the difference between Seedream 4.5 and Nano Banana Pro?
Seedream 4.5 vs Nano Banana Pro
What is the difference between Seedream 4.5 and Flux 2 Pro?
Seedream 4.5 vs Flux 2 Pro
What is the difference between Seedream 4.5 and GPT Image 1.5?
Seedream 4.5 vs GPT Image 1.5

Model comparison table

Not sure which model to choose? Compare the key differences between this model and its alternatives at a glance.

Workflow scope
Seedream 4.5
Text-to-Image and image editing
Seedream 5 Lite
Workflow scope — Seedream 5 Lite: A currently available alternative in the same category; open its detail page to compare capabilities and pricing.
Output setting
Seedream 4.5
basic 2K / high 4K
Seedream 5 Lite
Output setting — Seedream 5 Lite: A currently available alternative in the same category; open its detail page to compare capabilities and pricing.
Prompt limit
Seedream 4.5
3,000 characters
Seedream 5 Lite
Prompt limit — Seedream 5 Lite: A currently available alternative in the same category; open its detail page to compare capabilities and pricing.
Current pricing
Seedream 4.5
7 credits per request
Seedream 5 Lite
Current pricing — Seedream 5 Lite: A currently available alternative in the same category; open its detail page to compare capabilities and pricing.
When to compare it
Seedream 4.5
Test when 2K/4K output or multi-image editing matters
Seedream 5 Lite
When to compare it — Seedream 5 Lite: A currently available alternative in the same category; open its detail page to compare capabilities and pricing.

Real-world experience

Practical guidance based on public sources, current on-site limits, and representative tasks.

Quality

Seedream 4.5 / Quality — Keep the prompt and references fixed, then run `basic` and `high`. Compare detail, text, subject preservation, and usable dimensions. Both settings cost 7 credits, and the 4K label alone does not establish content quality.

Official showcases establish direction, not an outside guarantee

Seedream 4.5 / Official showcases establish direction, not an outside guarantee — Business usability still depends on your materials, copy, and review criteria.

Capabilities

Capabilities

Capabilities — ByteDance's official Seedream 4.5 material highlights multi-image editing, preservation of reference details, layout, and text rendering. The Kyeo workflow has its own concrete limits: prompts can contain up to 3,000 characters, 8 aspect ratios are available, and editing accepts up to 14 JPG, PNG, or WebP images at 10 MB each.

8 ratios and two output settings

8 ratios and two output settings — The workbench offers 1:1, 4:3, 3:4, 16:9, 9:16, 2:3, 3:2, and 21:9, with basic 2K or high 4K output. Each request currently costs 7 credits.

Prompt and reference limits

Prompt and reference limits — Prompts can contain up to 3,000 characters. Editing accepts up to 14 JPG, PNG, or WebP references at 10 MB each; text-to-image generation requires no image upload.

Use cases

Multi-reference composition and editing

Combine character, product, wardrobe, or style references, while stating what each image contributes and which elements must remain unchanged.

Posters and text-led visuals

Create a set of images around one theme for a social content series or campaign visual system.

Product and campaign variants

Change backgrounds, materials, or composition from an existing visual, with human checks for product appearance, trademarks, and required information.

Prompt tips

List what to preserve and what to change

For each reference, explain its role and identify the subject, pose, color, material, or background details that must remain or may change.

Provide display copy verbatim

For posters or product images, specify the exact copy, language, hierarchy, and placement.

Begin with the fewest references you need

14 images is only the editing limit. Start with the minimum set required by the brief, then remove conflicting images or constraints if subjects become confused.

Why choose it

One Kyeo entry covers text-only generation and reference-based editing.
ByteDance's official page explicitly demonstrates multi-image editing, reference preservation, layout, and text rendering.
The integration offers 8 aspect ratios plus 2K and 4K output settings.
Fixed 7-credit pricing makes request-count estimates straightforward.

What to know first

Keep fixed samples for long-lived workflows so composition, text, and style consistency can be rechecked after version updates.
Generate one representative sample with the chosen settings and confirm that it meets the goal before expanding the batch.
For extremely stylized artwork, compare models from different families.
Generated text, product details, trademarks, and legally sensitive content require human review.

Adjustable settings

Quickly evaluate whether this model fits your business use case.

Aspect ratio
Optional
Parameter type: Select
Default: 1:1
1:1
4:3
3:4
16:9
9:16
2:3
8 options
Quality
Optional
Parameter type: Select
Default: basic
Standard (2K)
High quality (4K)

Credit cost

7 credits per request

Kyeo AI currently charges a fixed 7 credits per request, with the same price for basic and high. Budget for retries, revisions, and the share of outputs that pass review.

Credit-saving tip

Before generating images or videos in bulk, run an A/B test with the same input on this model and similar alternatives. Confirm the result first to avoid wasting credits.

Frequently asked questions

Related models

Still deciding? Take a quick look at these similar alternatives.

Nano Banana Pro

Nano Banana Pro supports text-to-image generation and reference-image editing on Kyeo AI, with 1K, 2K, or 4K output. Each 1K or 2K request costs 8 credits, while 4K costs 14. Resolution is an output-size tier, not a guarantee of quality, accuracy, or detail.

AI image generation and editing model
8–14 credits per request
View detailsUse now

Flux 2 Pro

Flux 2 Pro / Capabilities — Flux 2 Pro — Text-to-image and editing use the same resolution pricing: 5 credits for 1K or 7 credits for 2K. 3–5,000 characters; Up to 8 images.

AI image model
5–7 credits per request
View detailsUse now

GPT Image 1.5

GPT Image 1.5 on Kyeo supports text-to-image and image-to-image with three aspect ratios and medium/high quality. Image editing accepts up to 16 JPG, PNG, or WebP references at 10 MB each. Medium costs 4 credits and high costs 22 credits. GPT Image 1.5 is OpenAI's deprecated previous image model. Kyeo retains generation and editing access, but that availability does not make it OpenAI's current default choice.

AI image model
4–22 credits per request
View detailsUse now

Sources

This page is based on vendor materials, model-family documentation, and Kyeo AI integration settings.

Source note

ByteDance's official Seedream 4.5 material highlights multi-image editing, preservation of reference details, layout, and text rendering. The Kyeo workflow has its own concrete limits: prompts can contain up to 3,000 characters, 8 aspect ratios are available, and editing accepts up to 14 JPG, PNG, or WebP images at 10 MB each.

Last updated: 2026-08-26
ByteDance Seed: Seedream 4.5
Official
Open original source
Volcengine: Seedream image generation and editing
Official
Open original source