Form OCR API that reads every box, typed or handwritten

Yes. Many forms are still filled in by hand, so Mindee is built to read handwriting as well as printed text, including capital letters, cursive and numbers written in boxes. It also detects whether checkboxes and radio buttons are ticked and returns them as true or false values. Handwriting quality varies a lot, so each field comes with a confidence score. Clearly written fields go through automatically, and hard-to-read entries are sent to a person instead of being guessed.

No. Traditional form OCR relies on templates that define where each field is on the page, so every new form or new version of an existing form requires a new template. Mindee works differently: you describe the fields you want to extract (for example, applicant name, policy number, date of incident, signature present), and the model finds them on the page, whatever their position. A redesigned form, a different version from a partner, or a form scanned at an angle doesn’t break your workflow.

Scan the forms in batches and send the files to the Mindee API, which processes documents in parallel and returns structured data in seconds per page. Records with high confidence scores can be loaded directly into your database or business application, and only uncertain fields need to be reviewed by a person. Teams can then turn archives of thousands of forms into searchable data in days rather than months, and handle new incoming forms through the same automated process.

Yes, your data is highly secured and protected during the entire conversion process. As an enterprise-grade OCR solution, Mindee takes data privacy very seriously. Here are the key security measures in place:

• Certified compliance: Mindee's infrastructure is SOC 2 Type II certified and fully GDPR compliant, ensuring strict industry standards for data protection.
• Encryption: All documents and extracted data are fully encrypted both in transit (during the API call) and at rest.
• Data localization: You have control over where your data is processed, with the option to host and process your data on either EU or US servers.
• Strict privacy: Mindee respects data confidentiality. Your processed documents are not shared with third parties, and your private data is not used to train global AI models.

In short, your sensitive documents are processed in a secure, isolated environment and remain completely under your control.

We provide a 14 days free trial so you can fully test any of our OCR models — no credit card required. After this, we offer different pricing tiers depending of the volume of pages processed and the features you might need. See the pricing page for more information.

The supported document formats are:

• PDF (application/pdf) – While technically a document format rather than an image, it is natively supported for all extraction models.
• JPEG / JPG (image/jpeg)
• PNG (image/png) – Must be non-animated.
• WebP (image/webp)
• TIFF / TIF (image/tiff) – Both single-page and multi-page TIFFs are supported and processed similarly to PDFs.
• HEIC (image/heic) – Apple's High-Efficiency Image Container format.

‍

📌 Important technical limitations to keep in mind:

To ensure the OCR pipeline runs smoothly and successfully returns your JSON, make sure your files adhere to the following constraints:

• Maximum file size: 100 MB per file.
• Maximum page count: Up to 200 pages per document.
• File state: The files cannot be encrypted, and PDFs must not be password-protected.