> ## Documentation Index
> Fetch the complete documentation index at: https://docs.vern.so/llms.txt
> Use this file to discover all available pages before exploring further.

# Export CSV

> Get a migration's validated data out as a clean CSV — addressed by template slug, downloaded or streamed straight into your own storage bucket.

```http theme={null}
GET https://app.vern.so/api/v1/migrations/{migration_id}/exports/{template_slug}.csv
```

Streams one sheet's **validated data** as a CSV, addressed by **template slug**.
This is synchronous — there's no run to poll. Columns come out in template order,
RFC 4180-escaped, and the body is streamed (gzip-compressed when your client
sends `Accept-Encoding: gzip`).

One call exports one sheet. To export a whole workbook, request each template's
slug in turn.

## Authentication

Requires an `x-api-key` header. See [Authentication](/migration-api/authentication).

## Path & query

<ParamField path="template_slug" type="string" required>
  The slug of the template whose sheet you want, with a `.csv` suffix — e.g.
  `contacts.csv`. [List your templates](/migration-api/list-templates) for slugs.
</ParamField>

<ParamField query="filter" type="string" default="valid">
  Which rows to include:

  * **`valid`** (default) — only fully clean rows. Any row with even one cell that
    failed validation is **excluded entirely**, so the count can be lower than the
    sheet's total.
  * **`all`** — every row, including those with invalid cells.
  * **`invalid`** — only the flagged rows (those with at least one invalid cell) —
    useful for surfacing what needs fixing.
</ParamField>

## Response

`200 OK` with `Content-Type: text/csv; charset=utf-8` and a `Content-Disposition`
attachment header. The body is streamed. An `X-Total-Records` header reports how
many records match the filter. When the request advertises gzip, the response is
gzip-encoded (`Content-Encoding: gzip`).

## Errors

| Status | Meaning                                                                                                |
| ------ | ------------------------------------------------------------------------------------------------------ |
| `400`  | The path doesn't end in `.csv`, or `filter` isn't one of `valid` / `all` / `invalid`.                  |
| `401`  | API key missing or invalid.                                                                            |
| `404`  | No template with that slug, no sheet for it in this migration, or the migration isn't in your account. |
| `429`  | Rate limit hit — back off and retry.                                                                   |
| `500`  | Server error.                                                                                          |

## Example

```bash theme={null}
# valid rows only (default)
curl "https://app.vern.so/api/v1/migrations/c0a8012e-.../exports/contacts.csv" \
  -H "x-api-key: $VERN_API_KEY" \
  -o contacts-clean.csv

# include rows with invalid cells
curl "https://app.vern.so/api/v1/migrations/c0a8012e-.../exports/contacts.csv?filter=all" \
  -H "x-api-key: $VERN_API_KEY" \
  -o contacts-all.csv

# only the flagged rows
curl "https://app.vern.so/api/v1/migrations/c0a8012e-.../exports/contacts.csv?filter=invalid" \
  -H "x-api-key: $VERN_API_KEY" \
  -o contacts-flagged.csv
```

## Into your own storage

The endpoint hands you the file in the HTTP response — Vern doesn't push it
anywhere. To land the clean CSV in **your own bucket**, pull it from your backend
and pipe it straight into object storage. The body streams, so you never have to
buffer the whole file:

```bash theme={null}
# Amazon S3
curl -s "https://app.vern.so/api/v1/migrations/c0a8012e-.../exports/contacts.csv" \
  -H "x-api-key: $VERN_API_KEY" \
  | aws s3 cp - s3://your-bucket/acme/contacts.csv

# Google Cloud Storage
curl -s "https://app.vern.so/api/v1/migrations/c0a8012e-.../exports/contacts.csv" \
  -H "x-api-key: $VERN_API_KEY" \
  | gcloud storage cp - gs://your-bucket/acme/contacts.csv
```

Any S3-compatible store (R2, MinIO, Spaces) works the same way. Keep your
`x-api-key` server-side — run this from your backend, never the browser. Vern
never receives your storage credentials; your backend relays the bytes.

<Note>
  Prefer Vern to deliver **directly** into a bucket you own — via a presigned URL
  or stored bucket credentials — instead of relaying it yourself? That's not
  available today. Tell us at [vish@vern.so](mailto:vish@vern.so) if you'd use it.
</Note>

## Next

* [Poll a run](/migration-api/poll-a-run) — confirm the execute completed before exporting.
* [Errors & limits](/migration-api/errors) — shared conventions.