As of: 2026-10-25 23:49 CET.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/m/seedream-5-0-pro/

# Seedream 5.0 Pro

ByteDance Seed, Seedream family. Status: current. Type: image generation. Released: 2026-07-08.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| BytePlus ModelArk |  |  | https://docs.byteplus.com/en/docs/ModelArk/2582774 |
| OpenRouter | `bytedance-seed/seedream-5-0-pro` | https://openrouter.ai/api/v1/images | https://openrouter.ai/bytedance-seed/seedream-5-0-pro |
| fal | `bytedance/seedream/v5/pro/text-to-image` |  | https://fal.ai/models/bytedance/seedream/v5/pro/text-to-image |
| fal (edit) | `bytedance/seedream/v5/pro/edit` |  |  |

## Pricing

- Per image: $0.045
- Per image high resolution: $0.09
- Per input image: $0.003
- Unit as published: USD per output image (standard / high_resolution), plus per input reference image; OpenRouter listing with provider Seed
- Source: https://openrouter.ai/api/v1/images/models/bytedance-seed/seedream-5-0-pro/endpoints
- Last checked: 2026-10-25

## Specs

- Input: text, image
- Output: image
- Context window: not published
- Max output: not published
- Open weights: no
- Licence: proprietary

## Where it stands

- Newer image generation models in the registry since its release: 11. The 5 newest:
  - FLUX 3 Image Klein (2026-10-15) https://postcutoff.com/m/flux-3-image-klein/
  - Krea 3 (Fast / High / Ultra modes) (2026-10-15) https://postcutoff.com/m/krea-3/
  - Nano Banana 2.1 (Gemini Nano Banana 2.1) (2026-10-06) https://postcutoff.com/m/gemini-nano-banana-2-1/
  - Qwen-Image-2.1 (2026-09-20) https://postcutoff.com/m/qwen-image-2-1/
  - GPT Image 2.5 Flare (2026-09-08) https://postcutoff.com/m/gpt-image-2-5-flare/
- Closed weights; access is hosted only.

## What stands out

- Region-precise editing with drawn boxes: Draw coloured boxes on the input image with one instruction per box; changes one element while keeping the rest of the frame intact; sketch completion. Source: https://fal.ai/learn/tools/how-to-use-seedream-5-0-pro-v2
- Layer separation: Splits an image into separate layers for production assets (layer count not stated by fal; secondary sources say 1 base plus up to 16 layers, unverified). Source: https://natlawreview.com/press-releases/fal-launches-seedream-50-pro-api-bringing-bytedances-new-image-model
- Native text in 14 languages and dense layouts: Native text rendering in 14 languages, control over dense layouts, up to 2K output with aspect ratios from 1:16 to 16:1. Source: https://natlawreview.com/press-releases/fal-launches-seedream-50-pro-api-bringing-bytedances-new-image-model
- Many reference images: Up to 10 reference images on fal's edit endpoint; OpenRouter's Seed endpoint accepts 0-14 input references, 1K/2K output and aspect ratios up to 21:9 plus auto. Source: https://openrouter.ai/api/v1/images/models/bytedance-seed/seedream-5-0-pro/endpoints

## Notes

Artificial Analysis text-to-image leaderboard (read 2026-10-25): #16, Elo 1100, listed release Jul 2026, $90.0 per 1k images (https://artificialanalysis.ai/text-to-image/arena/leaderboard-text); AA's $90 matches OpenRouter's high_resolution price, $45 is the standard tier. AA image-editing: tied 7-13, Elo 1131. API available from 2026-07-08 (fal press release dated 2026-07-09; Chinese press reprints of a Volcano Engine announcement give the same date, not read directly); third-party pages say it was unveiled at Volcano Engine FORCE 2026 on 2026-06-23 (unverified). Not verified: the official BytePlus/Volcengine model id (aggregators give dola-seedream-5-0-pro-260628 / doubao-seedream-5-0-pro-260628; BytePlus docs render only in a browser), official BytePlus pricing, and the exact output resolution list. fal prices: $0.0675 per image up to 1536x1536, $0.135 up to 2048x2048, edit +$0.0045 per extra reference (press release; fal's model page calls them tentative); fal edit endpoint bytedance/seedream/v5/pro/edit; batches up to 6 images per request on fal. OpenRouter versioned id bytedance-seed/seedream-5-0-pro-20260812. Siblings on OpenRouter: seedream-5-0-lite, seedream-5-0-flash. No timeline entry yet.

ByteDance's flagship image model of the Seedream 5.0 family, aimed at production assets (editing, layers, typography).

```bash
curl https://openrouter.ai/api/v1/images -H "Authorization: Bearer $OPENROUTER_API_KEY" -H "Content-Type: application/json" \
 -d '{"model":"bytedance-seed/seedream-5-0-pro","prompt":"A bilingual poster for a tea shop","resolution":"2K","aspect_ratio":"3:4"}'
```

Sources:
- https://openrouter.ai/bytedance-seed/seedream-5-0-pro
- https://openrouter.ai/api/v1/images/models/bytedance-seed/seedream-5-0-pro/endpoints
- https://docs.byteplus.com/en/docs/ModelArk/2582774
- https://fal.ai/models/bytedance/seedream/v5/pro/text-to-image
- https://fal.ai/learn/tools/how-to-use-seedream-5-0-pro-v2
- https://natlawreview.com/press-releases/fal-launches-seedream-50-pro-api-bringing-bytedances-new-image-model

Model page: https://postcutoff.com/m/seedream-5-0-pro/