PARSER GUIDES

API Access

Integrations

Access your data through the API

You can upload documents, read extracted data, and manage PARSER resources using the REST API. This lets you send documents programmatically (e.g. from your own system or scripts) instead of using the web interface. You need an API key from your team's API Keys page and a parser ID or classifier ID.

Each API key has its own permissions. It can apply to every parser or only selected parsers and classifiers. The key can perform only the actions and access only the resources chosen when it was created.

If you want an AI assistant to work with PARSER, use the MCP Server instead. MCP uses OAuth and lets agents call PARSER tools through natural language.

Where to find API keys

1. Open your team menu.

2. Click API Keys.

3. Create an API key.

4. Choose its permissions and whether it can access every parser or only selected parsers and classifiers.

5. Copy the key. Keep it secret; don't share it or commit it to code.

You can also reach Manage API Keys from Settings -> Integrations -> API Integration.

What you can do with the API

The exact actions available depend on the key's permissions and parser access. A key can be allowed to:

• Upload files - Send document files (PDF, images, Office, etc.) to an allowed parser or document classifier.
• Process from URLs - Send document URLs instead of files. PARSER downloads and processes them.
• List and manage parsers - Create parsers, update safe parser settings, and duplicate parser setup.
• List and manage classifiers - Create classifiers, update routing rules, inspect classifier documents, and reprocess skipped or failed classifier documents.
• Manage Fields & Tables - Add, update, delete, and reorder scalar fields and table fields.
• Search and manage documents - List documents, filter by status or folder, search extracted data, rename documents, update custom data, and move documents.
• Route uploads to a folder - Send uploads directly into a specific folder so your API intake stays organized.
• Attach external IDs and custom data - Add your own tracking ID and extra metadata so documents map cleanly to your internal system.
• Get extracted data - Fetch the extracted data for a processed document by document ID or external ID.
• Reprocess documents - Re-run extraction on an existing document without re-uploading (e.g. after adding new parser fields).
• Delete documents - Delete documents so you can re-upload with the same external ID.
• Use Reviews & Approvals - List reviewers, configure review stages, assign documents, and approve or reject documents.
• Export data - Export selected or filtered documents as JSON, XML, CSV, or XLSX.
• Manage webhook endpoints - Add, update, list, and delete parser webhook URLs.

Details, request format, and authentication (using your API key) are in the API Reference. Use the Authorization header with your API key when calling the API.

How to use it

1. Create an API key (see above).

2. Get your parser ID or classifier ID - Parser IDs are in Parser Settings -> General Settings. Classifier IDs are in Classifier Settings -> General.

3. Use the Process API to upload files or URLs. Send a parserId when you want one parser to extract data. Send a classifierId when you want PARSER to route mixed documents first.

4. Use parser document and export endpoints for extracted data. Use classifier document endpoints to inspect routed, not routed, or failed classifier documents.

The API uses your API key for authentication. Include it in the Authorization header when making requests.

Helpful upload options

When uploading through the API, you can also:

• Set external IDs so each document keeps your own reference number.
• Add custom data for extra context (for example, client code or source system).
• Pass a folder ID for parser uploads so documents land in the right folder automatically. Folder IDs are not supported for classifier uploads.

For new integrations, use:

• POST /api/v1/process/files with exactly one of parserId or classifierId
• POST /api/v1/process/urls with exactly one of parserId or classifierId

The older /api/v1/process/{parserId}/... endpoints still work for parser uploads.

These options are useful when you connect PARSER to your existing workflow tools.

AI agents and MCP

Use MCP when you want to ask an AI assistant to perform PARSER tasks, such as "list my parsers," "process this invoice," "find failed documents," or "export June invoices as CSV."

MCP is different from the REST API:

• REST API uses API keys and is best for your backend systems, scripts, and no-code tools.
• MCP uses OAuth and is best for AI clients that support remote MCP servers.

See MCP Server for setup steps, available tools, and troubleshooting.

Tips

• API keys belong to a team, but each key can be limited to selected parsers, classifiers, and actions.
• Keep keys secret - Don't share them or include them in public code or repositories.
• Rotate keys periodically for security (create a new one, update your system, then remove the old key).
• For most users, the web interface or email processing is easier than using the API. Use the API only if you need to integrate with your own systems or automate uploads.

Next steps

• API Reference - Endpoints, request/response format, and auth
• Document Classifiers - Route mixed batches before extraction

• MCP Server - Connect AI agents to PARSER

• Processing from URLs - Use the API to process documents from URLs
• Organizing with Folders - Keep API uploads sorted by client or workflow
• Integrations Overview - Email, Google Sheets, Zapier, Make, and webhooks