Seedream 5 Lite
prospolabs/seedream-5-lite
ByteDance's reasoning-based image model with chain-of-thought before generation. Generate from text or compose up to 10 reference images per edit. $0.014 per image — the cheapest reasoning image model on the web.
Starts from
Retail $0.035
Call the Seedream 5 Lite API on Prospolabs with a single POST to /v1/generate using model id seedream-5-lite and a mode field (one of: generate, edit). The API and the playground share the same USD price — $0.014 per image — 60% off market (cheapest reasoning image model on the web). Each call returns a generation_id and an output_url valid for 7 days, and failed runs are auto-refunded.
Quickstart
Send a POST request to /v1/generate with your API key. The model id is seedream-5-lite and a mode field selecting which input shape you're sending..
Create an image from a text prompt.
curl -X POST "https://api.prospolabs.com/v1/generate" \
-H "Authorization: Bearer $PROSPOLABS_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "seedream-5-lite",
"mode": "generate",
"prompt": "Minimalist product photo of a ceramic coffee mug, soft light",
"aspect_ratio": "16:9",
"size": "2K",
"num_images": 1,
"seed": 0
}'Backed by fal-ai/bytedance/seedream/v5/lite/text-to-image.
Supported parameters for Generate
| Parameter | Type | Default | Notes |
|---|---|---|---|
| mode | string | generate | Required. One of: generate, edit. |
| prompt | string | — | Required. |
| aspect_ratio | string | 16:9 | One of: auto, 1:1, 4:3, 3:4, 16:9, 9:16, 3:2, 2:3, 21:9. |
| size | string | 2K | One of: 2K, 3K. |
| num_images | number | 1 | Range: 1–6. |
| seed | number | — |
Response
{
"generation_id": "gen_abc123",
"status": "completed",
"model": "seedream-5-lite",
"mode": "generate",
"estimated_cost_usd": 0.60,
"output_url": "https://cdn.prospolabs.com/...",
"duration_ms": 2143
}Output retention: output_url is valid for 7 days after generation. After that, requests return HTTP 410 Gone with { "code": "output_expired" }. Download anything you want to keep.