> ## Documentation Index > Fetch the complete documentation index at: https://docs.linkup.so/llms.txt > Use this file to discover all available pages before exploring further. # Extract overview > Closed-beta agent that turns a web page into structured rows of data, delivered as a downloadable NDJSON file. **Extract** is in **closed beta**. Access is limited to a small group of users while we refine it — [request access](mailto:contact@linkup.so?subject=Request%20access%20to%20%2Fextract%20beta). Parameters, behavior, and response shape may change. **Extract** is Linkup's web data agent. Given a seed `url` and a natural-language description of the rows you want, it extracts the matching records and returns them as structured rows. Use cases include: * turning a listing page (team, catalog, directory) into a table of records, * pulling repeated entities (people, products, jobs) with consistent fields, and * collecting links from a page and verifying they resolve. The agent starts from the `url` you provide, extracts the rows described by `q`, and writes the result to a downloadable NDJSON file. ## Parameters | Parameter | Type | Default | Description | | ------------ | --------- | ------------ | ----------------------------------------------------------------------------------------------- | | `q` | `string` | *(required)* | Natural-language query describing which rows to extract and what each row should contain. | | `url` | `string` | *(required)* | The seed URL the extract task starts from. | | `schema` | `object` | `null` | JSON schema describing a single extracted row. When provided, every returned row must match it. | | `verifyUrls` | `boolean` | `false` | Whether URLs found in extracted rows are checked for reachability after extraction. | ### The `schema` parameter When `schema` is omitted, the agent infers the shape of each row from `q` and from the data structure on the website. Pass an explicit `schema` to pin the fields and types of every row. The schema describes **one row** — the agent applies it to each record it extracts. For example, to extract one row per team member: ```json theme={"system"} { "type": "object", "properties": { "name": { "type": "string", "description": "Full name" }, "role": { "type": "string" }, "profileUrl": { "type": "string", "format": "uri" } }, "required": ["name", "profileUrl"] } ``` ### The `verifyUrls` parameter When `verifyUrls` is `true`, any URL found in an extracted row is checked for reachability after extraction. This adds latency but filters out dead links. Leave it `false` when you only need the raw extracted values. ## Output Extract does **not** return the extracted rows in the API response. The `output` object contains a `resultUrl` — a download link to a separate file. You must make a second request to that URL to get the data. The link is valid for **24 hours**. Once the task is `"completed"`, `output` holds a link to a newline-delimited JSON (NDJSON) file — one extracted row per line — rather than the rows themselves: | Field | Type | Description | | -------------- | --------- | ---------------------------------------------------------------------------------------------------------------------------- | | `creditsUsed` | `number` | Credits used by this extract task. | | `resultUrl` | `string` | Download link for the extracted rows as an NDJSON file. **The rows live here, not in the API response.** Valid for 24 hours. | | `rowsReturned` | `integer` | Number of rows in the downloadable file. | To get the data: poll `GET /v1/extract/:id` until `status` is `"completed"`, read `output.resultUrl`, then make a separate `GET` request to that URL to download and parse the NDJSON file (one JSON object per line). ## Pricing **Extract** uses variable pricing: cost scales with crawl complexity — page size, number of rows extracted, pagination depth, and whether `verifyUrls` is enabled. Most completed tasks fall in the **\$2–10** range. The exact charge appears as `creditsUsed` in the task `output` once `status` is `"completed"`. No credit is deducted for failed tasks. A minimum account balance of **\$10** is required to submit an extract task. See [pricing](/pages/documentation/platform/pricing) for billing details. ## Async lifecycle `POST /v1/extract` returns immediately with a task identifier and `status` set to `"pending"`. Subsequent calls to `GET /v1/extract/:id` return the current state until `status` is `"completed"` or `"failed"`. `GET /v1/extract` is also available to list all extract tasks for the account. See the [list reference](/pages/documentation/endpoints/extract/list). ```text theme={"system"} POST /extract GET /extract/:id GET /extract/:id │ │ │ ▼ ▼ ▼ { id, status: { id, status: "processing" } { id, status: "completed", "pending" } (poll) output: { resultUrl } } ``` Poll at 30 seconds for long-running tasks. Polling above 1 request per second will be rate-limited. ## Example Create a Linkup account for free to get your API key. ```shell curl theme={"system"} # 1. Submit curl -X POST "https://api.linkup.so/v1/extract" \ -H "Authorization: Bearer $LINKUP_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "q": "All engineering team members with their name, role, and profile page", "url": "https://example.com/team", "schema": { "type": "object", "properties": { "name": { "type": "string", "description": "Full name" }, "role": { "type": "string" }, "profileUrl": { "type": "string", "format": "uri" } }, "required": ["name", "profileUrl"] }, "verifyUrls": true }' # → { "id": "01234-abcd-56789", "status": "pending", ... } # 2. Poll curl "https://api.linkup.so/v1/extract/01234-abcd-56789" \ -H "Authorization: Bearer $LINKUP_API_KEY" # 3. Download the rows once status is "completed" curl "" ``` `POST /v1/extract` returns the task envelope immediately, with `status` set to `"pending"` and `output` set to `null`. `GET /v1/extract/{id}` returns the same envelope; once `status` is `"completed"`, `output` is populated: ```json theme={"system"} { "id": "01234-abcd-56789", "type": "extract", "status": "completed", "createdAt": "2026-01-01T00:00:00.000Z", "updatedAt": "2026-01-01T00:03:12.000Z", "error": null, "input": { "q": "All engineering team members with their name, role, and profile page", "url": "https://example.com/team", "schema": { "type": "object", "properties": { "name": { "type": "string", "description": "Full name" }, "role": { "type": "string" }, "profileUrl": { "type": "string", "format": "uri" } }, "required": ["name", "profileUrl"] }, "verifyUrls": true }, "output": { "creditsUsed": 2.84, "resultUrl": "https:///extract/01234-abcd-56789.ndjson?...", "rowsReturned": 24 } } ``` Each line of the downloaded NDJSON file is one row matching `schema` (or the shape inferred from `q` and the data structure on the website): ```text theme={"system"} {"name": "Ada Lovelace", "role": "Staff Engineer", "profileUrl": "https://example.com/team/ada"} {"name": "Alan Turing", "role": "Principal Engineer", "profileUrl": "https://example.com/team/alan"} ``` ## Next Query phrasing, schema design, and polling. Tool definition and integration prompt. Full parameter spec and response schema.