Nightjar cannot read text from my PDF
Short answer
The PDF is probably an image rather than text. Select some text in it — if you cannot, there is no text layer and the model is reading it as pictures.
Why this happens
PDFs either carry a text layer or are scanned images in a PDF wrapper. Nightjar extracts the text layer when there is one. When there is not, pages are handled as images by a vision model, which is slower and less exact.
What to do
- 01Try selecting a sentence in the PDF. If nothing highlights, it is a scan.
- 02For a scanned file, make sure you are on a model that reads images — Qwen 3.5 or Gemma 4. Llama 3.2 1B cannot.
- 03Only the first 20 pages of a scanned PDF are processed. Split a longer document and ask about the part you need.
- 04If the original is available as a text PDF or a Word file, use that instead. It will be faster and more accurate.
Other ways people ask this
- nightjar pdf text not read
- Nightjar Browser: pdf text not read. How do I fix it?