API Reference

Structured Sheets

Copy Markdown

Open in Claude

Open in ChatGPT

Open in Cursor

---

Copy Markdown

View as Markdown

Tables

Convert uploaded spreadsheets into structured data. Creates relational tables from messy spreadsheet data.

List tables in structured sheet

$ deeptable structured-sheets:tables list

GET/v1/structured-sheets/{structured_sheet_id}/tables

Retrieve a table

$ deeptable structured-sheets:tables retrieve

GET/v1/structured-sheets/{structured_sheet_id}/tables/{table_id}

Download table data

$ deeptable structured-sheets:tables download

GET/v1/structured-sheets/{structured_sheet_id}/tables/{table_id}/download

ModelsExpand Collapse

table_response: object { id, created_at, name, 4 more }

Response representing a table extracted from a structured sheet.

This is returned from GET (retrieve) and list table endpoints. Table names use a composite format: {normalized_sheet_name}__{table_name}.

id: string

The unique identifier for this table.

created_at: string

The timestamp when this table was created.

name: string

Composite table name: {normalized_sheet_name}__{table_name}. Uses lowercase snake_case. Aggregation tables end with '__aggregations'. Two special metadata tables exist per structured sheet: '__deeptable_workbook_metadata' (workbook provenance info) and '__deeptable_table_overview' (summary of all tables). Example: 'staffing__head_count' or 'staffing__head_count__aggregations'.

object: "table"

The object type, which is always 'table'.

"table"

sheet_name: string

The original Excel sheet name this table came from.

structured_sheet_id: string

The ID of the structured sheet this table belongs to.

type: "relational" or "aggregation" or "tableless" or "metadata"

The type of table (relational, aggregation, tableless, or metadata).

"relational"

"aggregation"

"tableless"

"metadata"