All-in-One Extract

POST /v1/extract

Use this endpoint when you want a single request to fetch a page and return structured output.

Authentication

• External requests must include both X-API-Key and X-Project-Id
• Create or copy them from API key settings on the current site
• Use https://api.lexmount.com for COM projects and https://api.lexmount.cn for CN projects; credentials must match the API region

Request Body

{
  "extract": {
    "url": "https://example.com"
  }
}

Fields

Rules:

• At least one of extract.url or extract.dom_id is required
• Set trace.include_steps=true when you need workflow diagnostics

Extraction Options

{
  "extract": {
    "url": "https://example.com",
    "options": {
      "engine_preference": "chrome_cdp",
      "user_agent": "MyFetcher/1.0"
    }
  },
  "trace": { "include_steps": true }
}

When extract.dom_id is present, extraction reuses the snapshot instead of fetching the URL. Prefer sending only the returned dom_id rather than an unrelated URL. Changing the engine cannot re-render an existing snapshot; fetch the URL again for that.

/v1/extract currently uses a service-controlled 30-second extraction deadline. extract.options.timeout_ms does not override it, and the legacy workflow field does not configure it. The public gateway has its own timeout. Script/style filtering belongs to Dump DOM, not to an extraction output switch. template_id is no longer supported; templates are matched automatically. Do not rely on internal template-debug or force-generation fields. Unknown JSON fields are rejected.

Minimal Example

API_KEY='<your-api-key>'
PROJECT_ID='<your-project-id>'

curl -sS -X POST https://api.lexmount.cn/v1/extract \
  -H 'Content-Type: application/json' \
  -H "X-API-Key: $API_KEY" \
  -H "X-Project-Id: $PROJECT_ID" \
  -d '{"extract":{"url":"https://example.com"}}'

Response Shape

Top-level response fields:

Use request_id when you need to report a problem or correlate a client-side response with server-side request logs.

Common fields inside result:

• url
• final_url
• status_code
• title
• description
• main_text
• publish_time
• author
• language
• links
• images
• engine
• dom_id

Reuse an Existing dom_id

curl -sS -X POST https://api.lexmount.cn/v1/extract \
  -H 'Content-Type: application/json' \
  -H "X-API-Key: $API_KEY" \
  -H "X-Project-Id: $PROJECT_ID" \
  -d '{
    "extract": {
      "dom_id": "123"
    }
  }'

Enable Trace

By default, POST /v1/extract does not return trace.

To inspect workflow diagnostics, explicitly enable it:

curl -sS -X POST https://api.lexmount.cn/v1/extract \
  -H 'Content-Type: application/json' \
  -H "X-API-Key: $API_KEY" \
  -H "X-Project-Id: $PROJECT_ID" \
  -d '{
    "extract": {
      "url": "https://example.com"
    },
    "trace": {
      "include_steps": true
    }
  }'

Return Raw DOM

By default, raw_dom is not returned.

Enable it only when you actually need the captured DOM in the response:

curl -sS -X POST https://api.lexmount.cn/v1/extract \
  -H 'Content-Type: application/json' \
  -H "X-API-Key: $API_KEY" \
  -H "X-Project-Id: $PROJECT_ID" \
  -d '{
    "extract": {
      "url": "https://example.com"
    },
    "trace": {
      "include_raw_dom": true
    }
  }'

• raw_dom can be very large and is not recommended for normal production calls

Debug Mode

Turn on both debug fields only when you need to inspect the workflow and the captured page:

curl -sS -X POST https://api.lexmount.cn/v1/extract \
  -H 'Content-Type: application/json' \
  -H "X-API-Key: $API_KEY" \
  -H "X-Project-Id: $PROJECT_ID" \
  -d '{
    "extract": {
      "url": "https://example.com"
    },
    "trace": {
      "include_steps": true,
      "include_raw_dom": true
    }
  }'

• trace makes responses larger
• raw_dom can be very large and is not recommended for normal production calls

For common failures, see Common Errors.

Field Types and Optional Output

• result.status_code is the target page's status, not the API response's HTTP status.
• result.fetched_at is a timestamp string; publish_time, when present, is also a timestamp string.
• A links entry is { "text": "Example", "href": "https://example.com/" }; an images entry is { "src": "https://example.com/image.png", "alt": "Example" }.
• description, main_text, author, publish_time, language, links, images, debug, and extra may be omitted. Handle absent arrays before iterating them.
• extra may contain Open Graph, JSON-LD, or DOM summaries; its content varies by page and extraction method.
• metadata may contain dom_id, an opaque template_id, and server_elapsed_ms. The template identifier is not a supported input option.
• Even when requested, raw_dom may be omitted if no snapshot content is available. Individual trace steps also vary with the processing path.

See Dump DOM for snapshot reuse and recovery.