Seedream 5.0 Lite Text to Image API

bytedance/seedream/v5/lite/text-to-image
2K/3K · 8 aspect ratios · n 1–15

Seedream 5.0 Lite Text to Image transforms text prompts into 2K and 3K high-fidelity visuals, supporting multi-step visual reasoning, real-time web search enhancement, and batch generation of 1–15 images. It strictly adheres to spatial geometry and physical lighting while capturing time-sensitive details and structured infographics for production-ready creative assets.

Get API Key
Input
Required0/3000
Optional
Optional
Optional
OutputIdle

Generated images will appear here

Estimated cost: —
Continue using

Examples

text-to-image-01-output.png

A scientifically accurate educational infographic explaining the honeybee waggle dance, drawn in a warm naturalist illustration style on cream paper. A title across the top reads exactly "THE WAGGLE DANCE". Below it, four numbered panels in a 2x2 grid, each with one short caption in clean capital letters. Panel 1 caption "1. A FORAGER FINDS FLOWERS": a worker bee flies from a wooden hive to a field of purple lavender; a dotted flight line leads to the flowers, a small sun sits in the sky, and a thin arc marks a 40 degree angle between the sun direction and the flight line. Panel 2 caption "2. SHE DANCES ON THE COMB": a vertical honeycomb of hexagonal wax cells; the returning bee traces a figure-eight path whose straight waggle run is tilted 40 degrees to the right of vertical, with a dashed vertical reference line, a thin arc marking the 40 degree angle between that line and the waggle run, and a small upward arrow whose tip carries the single word "SUN" to show that straight up on the comb stands for the direction of the sun. Panel 3 caption "3. LONGER WAGGLE, FARTHER FOOD": three waggle runs of increasing length drawn side by side, each connected to a flower patch placed at an increasing distance from the hive. Panel 4 caption "4. SISTERS FOLLOW THE MAP": several bees that watched the dance fly out from the hive along the same 40 degree angle to the sun and reach the lavender field. The 40 degree angle must stay consistent in panels 1, 2 and 4. Anatomically correct bees with six legs and four wings, readable fine linework, every word spelled correctly, no other text, no watermark.

text-to-image-02-output.png

A cinematic documentary photograph inside a centuries-old glassblowing furnace room on the island of Murano, Venice, late at night. In the foreground, a master glassblower in a sweat-darkened grey linen shirt and a scorched leather apron turns a long blowpipe with a glowing orange gather of molten glass at its tip; his face is lit from below by the glow. In the middle ground, a young female assistant shields her face with a wooden paddle as she opens the round furnace door, revealing a white-hot interior that is the brightest light source in the room. In the background, partly hidden behind the assistant's shoulder, an older artisan sits at a steel bench shaping a translucent cobalt-blue vase with folded wet newspaper, a wisp of steam rising. Shelves along the soot-darkened brick wall hold finished clear, amber and green glass pieces that refract the orange light and cast colored caustics onto the bench. Correct occlusion and depth between the three people, every shadow falling away from the furnace, warm firelight contrasted with cool blue moonlight from one small window, visible heat shimmer above the furnace mouth, 35mm lens, natural skin texture, believable hands with five fingers, no text, no logos, no watermark.

text-to-image-03-output.png

An ultra-wide cinematic landscape photograph of the Altai Mountains in western Mongolia in deep winter at golden hour. In the left third of the foreground, a Kazakh eagle hunter sits on a sturdy brown Mongolian horse, wearing a fox-fur hat and a long embroidered wool coat; he raises his thick leather-gloved forearm as a golden eagle with fully spread wings lands on it, powder snow kicked up around the horse's hooves. In the middle distance on the right, a second rider leads a small caravan of four Bactrian camels across a frozen river, their long shadows stretching over the snow. Far behind, jagged snow-covered peaks glow pink and gold under a clear pale sky, each successive ridge lighter and hazier with distance. Correct relative scale between the eagle, the riders, the camels and the mountains, one consistent low sun from the right, visible breath vapor from the horse, fine detail in feathers and fur, no text, no watermark.

Seedream 5.0 Lite Text to Image

Seedream 5.0 Lite Text to Image is a next-generation lightweight, high-efficiency multimodal image generation model developed by the ByteDance Seed team. Advancing upon the Seedream family with enhanced understanding, reasoning, and synthesis, it natively delivers 2K and 3K resolution outputs across eight standard aspect ratios and custom dimensions. The model integrates real-time web retrieval to incorporate up-to-date real-world facts with chain-of-thought style multi-step visual reasoning. Whether producing topical campaigns driven by current trends, complex multi-subject scenes demanding coherent spatial perspective, or structured knowledge infographics, Seedream 5.0 Lite interprets instructions with exceptional depth. Supporting parallel batch generation of 1 to 15 images at 5 credits per image, it powers commercial creative pipelines across advertising, e-commerce, and digital media.

Why Choose Seedream 5.0 Lite Text to Image API?

  • Multi-Step Visual Reasoning & Spatial PerspectiveEmploys chain-of-thought visual inference to parse layered prompts, maintaining accurate physical scale, 3D depth of field, and natural occlusion across multi-subject compositions.

  • Real-Time Web Search for Timely Creative ContextDynamically retrieves online information during synthesis to reflect current events, breaking trends, and modern product details, overcoming static pre-training cutoff constraints.

  • Structured Infographics & Information VisualizationLeverages deep world knowledge and visual layout design to translate abstract technical workflows, data hierarchies, and educational concepts into structured, presentation-ready diagrams.

  • 2K/3K Native Outputs with 1–15 Batch GenerationProduces native 2K and 3K high-fidelity visuals directly with zero resolution surcharge, supporting concurrent generation of up to 15 distinct framing variations at 5 credits per image.

Parameters

ParameterRequirementDescription
promptRequired

Describe the image to generate in 3–3,000 characters. Leading and trailing whitespace is not counted.

sizeOptional

Output size: 2K, 3K, or an aspect ratio. For a custom size, enter WIDTHxHEIGHT (for example 2304x1728) or {"width":2304,"height":1728} with positive integers. Defaults to 1:1.

Default1:12K3K4:33:416:99:163:22:321:9
nOptional

Number of images to generate: integer 1–15. Defaults to 1.

Default123456789101112131415
enable_safety_checkerOptional

Set true to enable safety checking or false to turn it off. Defaults to true.

Defaulttrue

How to Use

  1. Compose Prompt and Visual DetailsDescribe the focal subject, artistic style, lighting, and composition in detail. Name specific real-world entities to trigger web knowledge retrieval when current context is needed.

  2. Select Resolution and Output CountChoose 2K or 3K resolution presets, select from eight aspect ratios or specify custom pixel dimensions, and set the image count n between 1 and 15.

  3. Submit Task and Retrieve ImagesSend your request to receive a unique task_id, poll the status endpoint until finished, or specify a callback_url to download generated high-resolution assets.

Pricing

Each image costs 5 credits ($0.025). The total is 5 × n credits, where n is the number of generated images.

UsageRateDetails
Standard generation5 credits/image · $0.025/image5 × n credits · $0.025 × n
n = 1 / 5 / 10 / 155 / 25 / 50 / 75 credits$0.025 / $0.125 / $0.250 / $0.375

Best Use Cases

  • Time-Sensitive Advertising & Topical CampaignsCreate trend-aware posters, social announcements, and marketing visuals that accurately incorporate real-time cultural events and product releases.

  • Educational Graphics & Conceptual InfographicsTurn complex curriculum topics, technical architectures, and business workflows into well-structured, visually appealing educational diagrams.

  • E-Commerce Products & Lifestyle PhotographyGenerate realistic studio environments, controlled reflections, and accurate surface textures for compelling product mockups and marketing assets.

  • Concept Art & Storyboard IdeationProduce up to 15 diverse perspectives, camera angles, and atmospheric treatments in a single run to accelerate pre-visualization and client pitching.

Pro Tips

  • Quote in-image text explicitly: Enclose any required typography in double quotation marks (for example, "SUMMER SALE" on the banner) and specify font weight, placement, and color palette.
  • Structure prompts for visual reasoning: Order descriptions by primary subject, relative spatial placement, surface textures, directional light sources, and overall environmental tone.
  • Leverage real-time entity names: Use exact commercial product names or topical keywords to allow the model's retrieval module to ground the output in accurate real-world context.
  • Batch explore then finalize: Submit an exploratory batch of n = 4 at default dimensions to survey composition options before fine-tuning the prompt for final high-resolution assets.

Notes

  • Text-only input mode: This endpoint accepts text prompts only. To edit existing images or blend multiple image references, use the Seedream 5.0 Lite Edit endpoint.
  • Prompt length requirements: Prompts must contain between 3 and 3,000 characters after trimming leading and trailing whitespace, with at least one non-whitespace character.
  • Automatic refund on task failure: Accounts are pre-deducted 5 × n credits upon task submission. If a task fails due to validation errors or network timeouts, pre-deducted credits are automatically refunded in full; partial generations refund the difference.

Related Models

Seedream 5.0 Lite Text to Image API FAQ

What is the Seedream 5.0 Lite Text to Image API?

Seedream 5.0 Lite Text to Image is a ByteDance Seed model for high-fidelity text-to-image generation. It turns written prompts into 2K and 3K high-resolution images with multi-step visual reasoning, real-time web retrieval, and structured information visualization, supporting concurrent batches of 1–15 images. Built on a unified multimodal architecture, it adheres strictly to spatial geometry and physical lighting while capturing time-sensitive details and complex layouts. You can call it programmatically or try it from the playground above.

How does Seedream 5.0 Lite Text to Image use real-time web retrieval?

During image synthesis, the model dynamically queries the live internet for up-to-date context, current events, emerging products, and cultural trends. When prompts reference recent real-world entities, it bypasses static training cutoffs to deliver factually grounded and contemporary visual elements.

What visual details does Seedream 5.0 Lite Text to Image improve through multi-step reasoning?

Multi-step visual reasoning provides chain-of-thought spatial analysis, improving 3D depth layering, object occlusion, scale proportions, and realistic light bouncing across materials. This prevents common generation defects such as floating objects, unnatural perspectives, or mismatched cast shadows.

Does Seedream 5.0 Lite Text to Image support infographics and knowledge diagrams?

Yes. The model incorporates comprehensive world knowledge and structural layout capabilities, allowing it to translate conceptual hierarchies, architectural blueprints, and workflows into clean, neatly aligned infographics, flowcharts, and presentation slides.

Can Seedream 5.0 Lite Text to Image generate multiple images in one batch?

Yes. You can generate between 1 and 15 distinct images in a single request by setting the integer n parameter. This allows you to explore diverse framing, lighting, and camera perspectives for the same prompt, billed only for successfully returned images.

What native resolutions and aspect ratios does Seedream 5.0 Lite Text to Image support?

It supports 2K and 3K native resolution presets, eight aspect ratios including 1:1, 4:3, 3:4, 16:9, 9:16, 3:2, 2:3, and 21:9, as well as custom dimensions via WIDTHxHEIGHT strings or width and height integer objects.

How are credits settled if a Seedream 5.0 Lite Text to Image task fails?

The platform enforces an automatic refund policy. Requests pre-deduct 5 × n credits upon submission. If generation fails due to input validation, timeouts, or network interruptions, all pre-deducted credits are automatically returned; partial batch completions refund the unused portion immediately.