How It Works :

1. Requirements and Optionals :

• A file in a supported media type, see below.
• A Talonic API key. Contact Talonic for details.
• Optional: A valid JSON schema. See JSON-Schema.org for instructions.
• Optional: A description of the data contained in the file; increases accuracy.

2. Submit a Request :

• Use the /process endpoint to submit a full job (extract + optional recommend + convert + optional validate). You can either upload a file or provide a URL to one, along with the JSON schema describing the expected results.
• Alternatively, use /extract to only extract markdown from the source without conversion, or /recommend to only generate a recommended JSON schema for the source.

‍

Sample cURL to submit a file directly :

curl -X PUT "https://api.talonic.ai/data-extractor/process" \

-H "Authorization: Bearer YOUR_API_KEY" \

-F "file=@/path/to/your/file.pdf" \

-F "json_schema={\"$schema\":\"http://json-schema.org/draft-07/schema#\",\"type\":\"object\",...}" \

-F "description=Optional description of the file"

‍

Sample cURL to submit a file URL :

curl -X PUT "https://api.talonic.ai/data-extractor/process" \

-H "Authorization: Bearer YOUR_API_KEY" \

-F "file_url=https://example.com/path/to/file.pdf" \

-F "json_schema={\"$schema\":\"http://json-schema.org/draft-07/schema#\",\"type\":\"object\",...}" \

-F "description=Optional description of the file"

‍

2. Submit a Request :

• To check the status and get the result of your processing job use the /process/{job_id} endpoint with the provided job_id.

‍