Scenario Seedream Images
Overview
Seedream, ByteDance's image family on Scenario, spans generation, reference-driven editing, and one
member that only takes images apart: Layerize splits a finished image into editable layers. The
members agree on little else, so discover them with search and read model_schema_get before
every run.
Connection and the core loop: see the scenario skill; model-agnostic image work: the
scenario-image skill. If a sibling skill named here is missing from your available skills, load it
with skill_get (Scenario MCP) or ask the user to install it
(https://github.com/scenario-labs/skills#install); unattended, proceed from tool schemas and flag
the gap.
Quick reference
At authoring time (fields and caps are per member, read the live schema):
| Member |
References |
Sizing |
Notes |
| 5.0 Pro |
referenceImages, up to 10 |
exact width and height, 672 to 3136 px, step 16 |
exact in-image text; 3500-char prompt |
| 5.0 Lite |
referenceImages, up to 14 |
width and height, up to 4K |
fast and cheap; 2048-char prompt |
| 5.0 Flash |
referenceImages, up to 10 |
exact width and height, 672 to 3136 px, step 16 |
cheapest; background keeps a reference's alpha |
| 4.5 |
referenceImages |
size (2K, 4K) plus aspectRatio enum |
subject-preserving edits; "auto" ratio follows the source |
| 5.0 Pro Layerize |
one image, required |
size tier: auto, 1K, 1.5K, 2K |
splits, never generates; prompt optional |
| 5.0 Flash Layerize |
one image, required |
size tier: auto, 1K, 1.5K, 2K |
same contract minus optimizePromptMode |
On the generators, mode follows from the inputs: empty referenceImages is text-to-image, one or
more is an edit or a multi-reference generation (with several, the prompt gives each reference a
role by position), and the array shape holds even for one asset. Sequence mode
(sequentialImageGeneration: "auto" plus maxImages) lets Lite and 4.5 return a related set in one
run, input plus generated capped at 15 images; Pro has no sequence fields. Price per image differs
widely: a 2048 px square quoted 4 CU on Flash, 6 on Lite, and 18 on Pro at authoring time, and Pro
took about two minutes against under one, so iterate on the cheap members and spend Pro on finals;
dry_run both before a batch (the estimate prices the run exactly as submitted, a whole sequence
included). Pro and Pro Layerize carry optimizePromptMode: the default standard reasons about the
prompt first and is slower; fast costs the same and is usually enough when a reference already
sets the composition (on Layerize it trades some split fidelity for speed). Flash's schema says
width and height apply only when Resolution is Custom, but Flash has no Resolution field: the
pixels apply directly.
Keeping a transparent asset transparent
Flash's background: "transparent" preserves alpha through an edit; it does not cut a subject out.
It keeps the transparent surround, not translucency inside the subject: glass at alpha 170 came back
near-opaque in one authoring-time test, so composite a translucent part in post. The field defaults
to opaque, so a restyle that omits it flattens a valid cutout: set it on every run that must stay
transparent. It applies only with exactly one reference image that already carries an alpha channel,
and is ignored otherwise: a text-only run, a flattened JPEG reference, or two references all come
back opaque without an error. So restyle or recolor an existing cutout (a sprite, an icon, a prop)
on Flash with the PNG as the one reference, and check the result's alpha on the downloaded file
before shipping it, since asset_display composites transparency away: with Pillow
(uv run --with pillow when it is missing), Image.open(path).getchannel("A").getextrema() should
start at 0 (fully transparent pixels exist), and an image with no A channel was flattened. To get
a cutout in the first place, generate on a plain field and run background removal (the
scenario-image-editing skill), or split the image with Layerize.
Write the exact copy into the prompt
Pro renders legible in-image text, multi-line layouts and non-Latin scripts included. Quote the
exact strings in the prompt instead of paraphrasing them, give each a position and a size rank
(headline across the top, date line small at the foot), and keep to a few elements at one or two
sizes: letters are drawn, not typeset, so paragraphs and many small labels turn to shapes, and copy
past that budget (tour dates, credits) is composited in post from the start. Proofread with
asset_display; a word that garbles is spelled out letter by letter after the quoted string on the
rerun, and one that still garbles goes to post. To swap copy on a reference, quote the new string,
pin its position, and require the original typeface, size and color with everything else unchanged.
Layerize: one image in, an editable stack out
Layerize returns a base layer plus up to 16 transparent PNG cutouts, rebuilding the background
behind whatever it lifts. The prompt picks the mode: empty runs a full automatic split; an
enumerated list of parts ("Separate this poster into transparent layers: headline, product, shadow,
background") cuts better than "all layers"; <bbox>x1 y1 x2 y2</bbox> on a 0 to 1000 grid, origin
top left, confines the split to one region. Set size explicitly, since auto inherits the source's
tier and the tier moves the price; cost is otherwise flat per run, not per layer. Layers come back
cropped to their own bounds with bbox and z-index metadata, not aligned to the source canvas, and
there is no PSD export.
Worked example: a poster, then its layers
search with target="models", query="seedream", public=true. Prefer the newest
non-deprecated hit for the job, e.g. model_bytedance-seedream-5-0-pro (a live hit at authoring
time: re-discover each session).
model_schema_get with that id: sizing fields, caps, defaults.
model_run with that id, dry_run=true, and
parameters={"prompt": "Concert poster, teal and cream, screen-print grain. Headline \"MIDNIGHT ORBIT\" across the top, date line \"Nov 14, Union Hall\" small at the foot.", "width": 1600, "height": 2368};
both sizing fields move price.
- Rerun
model_run with wait=false, then jobs_wait with the returned job id, re-called with
pending_job_ids on timeout, never a second model_run.
asset_display and proofread the rendered text.
model_schema_get on the Layerize hit, then model_run with that id and
parameters={"image": "<poster asset id>", "prompt": "Separate this poster into transparent layers: the headline text, the date line, and the background. Clean edges, complete transparency.", "size": "2K"}.
jobs_wait, then asset_display each layer and asset_download the keepers.
Common mistakes
- Reusing one member's parameter block on another: pixels on Pro and Lite, tier enums on 4.5 and
Layerize; pixels sent to an enum field are rejected.
- Passing
referenceImages to Layerize or image to the generators: the input field's name and
shape differ per member.
- Expecting Layerize layers to overlay the source directly: each is cropped to its own bounds, so
reposition with the returned metadata.
- Asking Pro for a sequence: the sequence fields existed on Lite and 4.5 only.
- Moving a 3000-character prompt from Pro to Lite: prompt caps are per member.
- Setting Flash's
background to transparent on a text-only run or a flattened reference: it is
silently ignored, and the image comes back opaque.
1---2name: scenario-seedream3description: "Use when generating or editing images with Seedream models on Scenario via MCP: text-to-image, image-to-image editing with reference images, posters or packaging with exact in-image text (non-Latin scripts too), subject-preserving edits, sets of related images in one run, editing a transparent asset without losing its alpha, or splitting an image into transparent PNG layers with Layerize. Keywords: Seedream 5.0 Pro, Lite, Flash, 4.5, Layerize, ByteDance, txt2img, img2img, layer extraction."4license: MIT5metadata:6 source: https://skillmd.com/skills/scenario-labs/scenario-seedream7---89# Scenario Seedream Images1011## Overview1213Seedream, ByteDance's image family on Scenario, spans generation, reference-driven editing, and one14member that only takes images apart: Layerize splits a finished image into editable layers. The15members agree on little else, so discover them with `search` and read `model_schema_get` before16every run.1718Connection and the core loop: see the `scenario` skill; model-agnostic image work: the19`scenario-image` skill. If a sibling skill named here is missing from your available skills, load it20with `skill_get` (Scenario MCP) or ask the user to install it21(https://github.com/scenario-labs/skills#install); unattended, proceed from tool schemas and flag22the gap.2324## Quick reference2526At authoring time (fields and caps are per member, read the live schema):2728| Member | References | Sizing | Notes |29| ------------------ | --------------------------- | --------------------------------------------------- | ----------------------------------------------------------- |30| 5.0 Pro | `referenceImages`, up to 10 | exact `width` and `height`, 672 to 3136 px, step 16 | exact in-image text; 3500-char prompt |31| 5.0 Lite | `referenceImages`, up to 14 | `width` and `height`, up to 4K | fast and cheap; 2048-char prompt |32| 5.0 Flash | `referenceImages`, up to 10 | exact `width` and `height`, 672 to 3136 px, step 16 | cheapest; `background` keeps a reference's alpha |33| 4.5 | `referenceImages` | `size` (2K, 4K) plus `aspectRatio` enum | subject-preserving edits; `"auto"` ratio follows the source |34| 5.0 Pro Layerize | one `image`, required | `size` tier: auto, 1K, 1.5K, 2K | splits, never generates; prompt optional |35| 5.0 Flash Layerize | one `image`, required | `size` tier: auto, 1K, 1.5K, 2K | same contract minus `optimizePromptMode` |3637On the generators, mode follows from the inputs: empty `referenceImages` is text-to-image, one or38more is an edit or a multi-reference generation (with several, the prompt gives each reference a39role by position), and the array shape holds even for one asset. Sequence mode40(`sequentialImageGeneration: "auto"` plus `maxImages`) lets Lite and 4.5 return a related set in one41run, input plus generated capped at 15 images; Pro has no sequence fields. Price per image differs42widely: a 2048 px square quoted 4 CU on Flash, 6 on Lite, and 18 on Pro at authoring time, and Pro43took about two minutes against under one, so iterate on the cheap members and spend Pro on finals;44`dry_run` both before a batch (the estimate prices the run exactly as submitted, a whole sequence45included). Pro and Pro Layerize carry `optimizePromptMode`: the default `standard` reasons about the46prompt first and is slower; `fast` costs the same and is usually enough when a reference already47sets the composition (on Layerize it trades some split fidelity for speed). Flash's schema says48`width` and `height` apply only when Resolution is Custom, but Flash has no Resolution field: the49pixels apply directly.5051## Keeping a transparent asset transparent5253Flash's `background: "transparent"` preserves alpha through an edit; it does not cut a subject out.54It keeps the transparent surround, not translucency inside the subject: glass at alpha 170 came back55near-opaque in one authoring-time test, so composite a translucent part in post. The field defaults56to `opaque`, so a restyle that omits it flattens a valid cutout: set it on every run that must stay57transparent. It applies only with exactly one reference image that already carries an alpha channel,58and is ignored otherwise: a text-only run, a flattened JPEG reference, or two references all come59back opaque without an error. So restyle or recolor an existing cutout (a sprite, an icon, a prop)60on Flash with the PNG as the one reference, and check the result's alpha on the downloaded file61before shipping it, since `asset_display` composites transparency away: with Pillow62(`uv run --with pillow` when it is missing), `Image.open(path).getchannel("A").getextrema()` should63start at 0 (fully transparent pixels exist), and an image with no `A` channel was flattened. To get64a cutout in the first place, generate on a plain field and run background removal (the65`scenario-image-editing` skill), or split the image with Layerize.6667## Write the exact copy into the prompt6869Pro renders legible in-image text, multi-line layouts and non-Latin scripts included. Quote the70exact strings in the prompt instead of paraphrasing them, give each a position and a size rank71(headline across the top, date line small at the foot), and keep to a few elements at one or two72sizes: letters are drawn, not typeset, so paragraphs and many small labels turn to shapes, and copy73past that budget (tour dates, credits) is composited in post from the start. Proofread with74`asset_display`; a word that garbles is spelled out letter by letter after the quoted string on the75rerun, and one that still garbles goes to post. To swap copy on a reference, quote the new string,76pin its position, and require the original typeface, size and color with everything else unchanged.7778## Layerize: one image in, an editable stack out7980Layerize returns a base layer plus up to 16 transparent PNG cutouts, rebuilding the background81behind whatever it lifts. The prompt picks the mode: empty runs a full automatic split; an82enumerated list of parts ("Separate this poster into transparent layers: headline, product, shadow,83background") cuts better than "all layers"; `<bbox>x1 y1 x2 y2</bbox>` on a 0 to 1000 grid, origin84top left, confines the split to one region. Set `size` explicitly, since auto inherits the source's85tier and the tier moves the price; cost is otherwise flat per run, not per layer. Layers come back86cropped to their own bounds with bbox and z-index metadata, not aligned to the source canvas, and87there is no PSD export.8889## Worked example: a poster, then its layers90911. `search` with `target="models"`, `query="seedream"`, `public=true`. Prefer the newest92 non-deprecated hit for the job, e.g. `model_bytedance-seedream-5-0-pro` (a live hit at authoring93 time: re-discover each session).942. `model_schema_get` with that id: sizing fields, caps, defaults.953. `model_run` with that id, `dry_run=true`, and96 `parameters={"prompt": "Concert poster, teal and cream, screen-print grain. Headline \"MIDNIGHT ORBIT\" across the top, date line \"Nov 14, Union Hall\" small at the foot.", "width": 1600, "height": 2368}`;97 both sizing fields move price.984. Rerun `model_run` with `wait=false`, then `jobs_wait` with the returned job id, re-called with99 `pending_job_ids` on timeout, never a second `model_run`.1005. `asset_display` and proofread the rendered text.1016. `model_schema_get` on the Layerize hit, then `model_run` with that id and102 `parameters={"image": "<poster asset id>", "prompt": "Separate this poster into transparent layers: the headline text, the date line, and the background. Clean edges, complete transparency.", "size": "2K"}`.1037. `jobs_wait`, then `asset_display` each layer and `asset_download` the keepers.104105## Common mistakes106107- Reusing one member's parameter block on another: pixels on Pro and Lite, tier enums on 4.5 and108 Layerize; pixels sent to an enum field are rejected.109- Passing `referenceImages` to Layerize or `image` to the generators: the input field's name and110 shape differ per member.111- Expecting Layerize layers to overlay the source directly: each is cropped to its own bounds, so112 reposition with the returned metadata.113- Asking Pro for a sequence: the sequence fields existed on Lite and 4.5 only.114- Moving a 3000-character prompt from Pro to Lite: prompt caps are per member.115- Setting Flash's `background` to `transparent` on a text-only run or a flattened reference: it is116 silently ignored, and the image comes back opaque.