Limits & Constraints
Supported File Types
We currently support .jpg, .png and .pdf format.
| Type | MIME Type | Extensions |
|---|---|---|
| Image JPEG (Joint Photographic Experts Group) | image/jpeg | .jpeg, .jpg |
| Image PNG (Portable Network Graphics) | image/png | .png |
| PDF (Adobe Portable Document Format) | application/pdf |
Input Size Limitations
- Maximum file size (images and PDFs): 4 MB.
- Maximum number of pages for PDF files (Sync Inference): 15 pages
- Maximum number of pages for PDF files (Async Inference): 50 pages
Note
Be advised that within our file processing workflow, pages from PDF documents are extracted as images and subject to the size limitations outlined above. See Working with PDFs for more informations.
Working with Images
When it comes to images, our APIs have a quicker upload and processing time.
Tip
If you want faster processing, then images are a better option compared to PDFs. However, using PDFs with metadata leads to more accurate results.
- Upscaling: do not upscale low-resolution images yourself, as this can reduce the accuracy of the algorithm. It is advisable to use higher-resolution images whenever possible.
- Do not preprocess images: you do not need to convert your image to black and white or adjust the brightness and contrast. Inappropriate preprocessing can negatively impact performance.
Working with PDFs
Our APIs support multi-page PDF files. However, the processing time may be longer compared to image processing, because we need to convert the PDF to image first.
Info
Now our standard is 150 DPI.
- Use native pdf: native PDFs are to process and generally deliver superior performance and accuracy than scanned PDFs.
- DPI: keep in mind this value if you work with our API and have to deal with bounding boxes visualization.