GET
Get a dataset summary

Authorizations

Authorization
string
header
required

Your Gwenlake API key, sent as Authorization: Bearer <key>.

Headers

x-user
string | null

Path Parameters

dataset_id
string
required

Response

Successful Response

Unified storage summary for a dataset, regardless of type.

  • iceberg: read from the table's current snapshot summary (exact, cheap). num_files is always None — the table's internal data-file count is an implementation detail, not a user-visible file count.
  • files: aggregated from the dataset's stored files; num_rows is counted over the Parquet files (None when none are present / countable).
  • vector: num_rows is the LanceDB table's row count, dimensions the vector width and model the embedding model; num_files/size_bytes walk the table's storage.

Fields are None when unknown (e.g. an Iceberg table with no data yet, or a file dataset with no Parquet).

type
string
required
object
string
default:dataset.summary
num_rows
integer | null
num_files
integer | null
size_bytes
integer | null
last_modified
string<date-time> | null
model
string | null
dimensions
integer | null