Seedream 5.0 Lite Text to Image API
/v1/tasks All models are called through the Unified Async API POST /v1/tasks endpoint; only the input fields differ (see input parameters below).
Model summary
| Model name | seedream-5.0-lite/text-to-image |
|---|---|
| Type | Image generation (text-to-image) |
| Endpoint | POST /v1/tasks |
| Pricing | See HiAPI Pricing |
ByteDance Seedream 5.0 Lite text-to-image: reasoning-guided generation with real-time web knowledge and accurate multilingual text rendering — ideal for posters, menus and infographics, with 2K and 4K at the same price.
Production guidance
- For production, pass callback.url at the top level of the request body so HiAPI can notify your service when the task reaches a terminal state.
- GET /v1/tasks/:id is better for local debugging, low-volume jobs, or fallback reconciliation if a callback is missed.
- Use callback.when=final. Both success and fail are terminal states, so your service should deduplicate by taskId.
Best suited for
Posters, menus and infographics are a strength, with accurate multilingual text rendering — just spell out every word you want in the prompt.
promptBuilt-in web retrieval and reasoning let it visualize current events and data, with good factual accuracy in annotated infographics.
promptThe 4K tier outputs roughly a 4096px long edge (verified) at the same price as 2K — pick 4K for print and large displays.
resolutionEight ratios from 1:1 to 21:9 adapt one prompt to feeds, banners and vertical posters.
aspect_ratioRequest parameters
model string required Fixed value seedream-5.0-lite/text-to-image.
input object required Business parameters. Put Seedream 5.0 Lite Text to Image-specific configuration here.
prompt string required Text description of the image, 3–3000 characters. The model reasons about intent and can pull in real-time web knowledge; spell out every word you want rendered in the image.
aspect_ratio enum required Width-to-height ratio of the output image.
resolution enum required Output resolution tier: 2K standard (≈2300px long edge) or 4K ultra (≈4096px long edge, verified). Both tiers cost the same.
callback object optional Optional callback configuration. When set, HiAPI notifies your service when the task reaches a terminal state.
url string required Required when callback is set; HTTPS URL that receives terminal task notifications.
when enum optional Callback trigger timing. Use final.
Example requests
3:4 / 2K — bilingual title poster that exercises text rendering and layout.
{
"model": "seedream-5.0-lite/text-to-image",
"input": {
"prompt": "A hand-painted museum poster for a bronze-age exhibition, featuring an ancient Chinese bronze ding vessel as the centerpiece, elegant Chinese calligraphy title 青铜时代 at the top with smaller English subtitle BRONZE AGE below it, cinematic composition, dramatic directional lighting, rich cohesive color grade of aged gold and deep teal, ultra-detailed, premium commercial finish, subtle film grain",
"aspect_ratio": "3:4",
"resolution": "2K"
}
}4:3 / 4K — eight planets with orbital-period callouts, exercising factual accuracy.
{
"model": "seedream-5.0-lite/text-to-image",
"input": {
"prompt": "A modern flat-design educational infographic poster of the solar system, minimalist geometric planets each labeled with clean sans-serif typography: Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, Neptune, thin white orbital rings on a deep space gradient from indigo to black, tasteful data callouts with orbital periods, museum print quality",
"aspect_ratio": "4:3",
"resolution": "4K"
}
}16:9 / 4K — three languages of neon signage, all legible.
{
"model": "seedream-5.0-lite/text-to-image",
"input": {
"prompt": "A rain-soaked neon street food alley at night in a cinematic Asian metropolis, glowing signs with clearly readable text: Chinese 夜宵小馆, English NIGHT NOODLES, Japanese ラーメン横丁, wet asphalt mirror reflections, steam rising from a food stall, cinematic composition, rich cohesive cyan and magenta color grade, ultra-detailed, subtle film grain",
"aspect_ratio": "16:9",
"resolution": "4K"
}
}Getting the result
- The response returns a taskId immediately without waiting for generation to finish.
- In production, prefer waiting for callback.url to receive the terminal notification. For local debugging, poll GET /v1/tasks/:id.
- When status=success, download the generated image from output[].url.
- When status=fail, fix the request based on the returned error instead of retrying the same invalid payload.
FAQ
What is the difference between 2K and 4K? Do they cost the same?
The 2K tier outputs roughly a 2300px long edge and 4K roughly 4096px (both verified). Both tiers cost the same — billing is per image regardless of resolution. See the pricing page for live rates. View pricing
Can it render text in the image accurately?
Multilingual text rendering is a core strength. Spell out every word you want in the prompt and ask for "accurate text" — avoid pinyin or translated names. Rare glyphs in stylized fonts can occasionally slip, so zoom in to check important copy.
What do web knowledge and reasoning mean? Any extra parameters?
The model has built-in real-time web retrieval and intent reasoning. Just describe what you need in the prompt (for example, today's weather or a set of real figures) — no extra parameters required.
How do I get the generated image?
The request returns a taskId immediately. When the task reaches a terminal state, download the image from output[].url. In production, pass callback.url at the top level of the request body to receive the terminal notification instead of polling.