Skip to main content
Request and response schemas for model inference. These Pydantic models define the JSON payloads exchanged between clients and a deployed model endpoint (e.g. SageMaker or a local server). Each request/response pair corresponds to a single tile.

​
ModelRequest

Payload sent to the model endpoint for a single tile. The tile image is base64-encoded for JSON transport. Bulk RNA counts can optionally be included for multimodal models.
​
str
required
Base64-encoded tile image (PNG).
​
list[float] | None
Optional bulk RNA counts vector.
​
str
required
Stem of the source slide filename.
​
int
required
Tile x-coordinate in the slide.
​
int
required
Tile y-coordinate in the slide.
​
int
required
Tile width in pixels.
​
int
required
Tile height in pixels.
​
float | None
Fraction of tissue in the tile.
​
int
required
Index of this tile in the extraction plan.

​
image

Decodes and returns the tile image as PIL.

​
ModelResponse

Payload returned by the model endpoint for a single tile. Carries the model output together with the full tile metadata from the originating request, so that downstream writers can persist everything without needing a separate schema.
​
list[float]
Model output vector for this tile.
​
str
Stem of the source slide filename.
​
int
Tile x-coordinate in the slide.
​
int
Tile y-coordinate in the slide.
​
int
Tile width in pixels.
​
int
Tile height in pixels.
​
float | None
Fraction of tissue in the tile.
​
int
Index of this tile in the extraction plan.
​
float | None
Extraction resolution (MPP) if available.

​
SageMakerRequest

Extended request for the SageMaker /invocations endpoint. Adds a model_name field so the server can dispatch to the correct model and a mode field to select the scheduler ("prediction" or "embedding").