> ## Documentation Index > Fetch the complete documentation index at: https://infino.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. # Append rows > Append rows to a table. The body is an Arrow IPC stream, or a JSON `{"data": [...]}` envelope. A column the table does not have is added, typed from the rows; a value that disagrees with a column it does have is rejected. ## OpenAPI ````yaml /api-reference/openapi.json post /v1/append/{database} openapi: 3.1.0 info: title: Infino API description: >- The Infino hosted data-plane API: per-database table operations — create, ingest, search, and SQL. version: 0.1.0 servers: - url: https://api.platform.infino.ws description: Infino Cloud security: - api_key: [] tags: - name: Databases description: Create, list, and delete the databases in your account. - name: Tables description: Create, drop, and list tables, and describe a table's schema. - name: Rows description: Append, update, and delete rows. - name: Search description: BM25, vector, and hybrid search, token and exact match, count, and SQL. paths: /v1/append/{database}: post: tags: - Rows summary: Append rows description: >- Append rows to a table. The body is an Arrow IPC stream, or a JSON `{"data": [...]}` envelope. A column the table does not have is added, typed from the rows; a value that disagrees with a column it does have is rejected. operationId: append parameters: - name: database in: path description: Target database. required: true schema: type: string - name: table in: query description: Target table. required: true schema: type: string requestBody: description: >- The rows to append: a JSON `{"data": [...]}` envelope of rows keyed by column name (`application/json`), or an Arrow IPC stream of one record batch (`application/vnd.apache.arrow.stream`). content: application/json: schema: $ref: '#/components/schemas/RowsEnvelope' application/vnd.apache.arrow.stream: schema: $ref: '#/components/schemas/ArrowIpcRows' required: true responses: '200': description: >- Rows appended. Body is an Arrow IPC row stream (application/vnd.apache.arrow.stream), or a JSON `{"data": [...]}` envelope. '400': description: >- Invalid request body, or rows the table refuses. A key the table has no column for adds one, typed from the values; a value whose type disagrees with a column the table already has is refused, naming the column, the type it holds and the one offered. Growing past the table's `max_fields` is refused the same way. A column a write adds is held to the same two rules a declared one is: its values may allocate at most 16384 bytes each even when null (a struct's fixed-width children add up, so `Struct{vector(4096), i32}` is 16388 and refused), and the engine must be able to write its type. Not transient: the same body fails identically. Change the value, or change the column with a `schema` patch first. '401': description: Missing or invalid API key. '403': description: >- The account already stores one of its limits — rows, or vectors (one per row of a table with a vector column) — so no more can be added. Which ceiling applies depends on the account: one with no payment method on file is held to the smaller free-plan ceiling, and adding a payment method raises it. The response body names which limit was reached, the numbers, and whether a payment method would raise it. Delete rows to free space; reads and deletes are never refused for this reason. Not transient: retrying without freeing space fails identically. '404': description: >- The target table does not exist. Append never creates a table — create it first with create_table. '409': description: >- Another write to the same table was in flight, so this one was not applied. A table takes one write at a time. Transient — reissue the identical request after the `Retry-After` interval, or batch more rows per request to need fewer of them. '503': description: >- The database's workers are still activating, or no capacity is free to place them. Transient — retry after the `Retry-After` interval. security: - api_key: [] components: schemas: RowsEnvelope: type: object description: >- JSON row envelope for `POST /v1/append/{database}` and `POST /v1/update/{database}` — the `Content-Type: application/json` alternative to an Arrow IPC stream body. Rows are JSON objects keyed by column name, converted server-side into one Arrow batch against the table's schema. A key the table has no column for **adds** one, typed from the values in the request and nullable, so earlier rows read back null for it. That is the table's behaviour and not this encoding's: an Arrow batch carrying a column the table lacks adds it on the same terms. What is specific to JSON is how a row is read — a nested object flattens to dot-path columns, an array becomes a list column, and the column's type is inferred from the values rather than stated by the caller. A value whose type disagrees with a column the table already has is a 400, naming the column, the type it holds and the one offered; the write lands no rows. Growing past the table's `max_fields` is a 400 too. Neither is retryable as sent: change the value, or change the column with a `schema` patch first. Appending never creates a table — an unknown table name is a 404. required: - data properties: data: type: array items: type: object description: Rows as JSON objects keyed by column name. additionalProperties: false ArrowIpcRows: type: string format: binary description: |- A raw Arrow IPC stream request body: one record batch of rows, sent as `Content-Type: application/vnd.apache.arrow.stream`. The most efficient ingest path — no JSON conversion; the batch's schema must match the table's. securitySchemes: api_key: type: http scheme: bearer ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.