Knowledge
Supported files and limits
File formats, size limits, and file statuses for company knowledge Collections.
This page lists the files company knowledge accepts, its limits, and what each file status means.
File formats
| Format | Extensions | Notes |
|---|---|---|
| Text | .txt, .text | Must be UTF-8 |
| Markdown | .md, .markdown | Must be UTF-8. Headings become sections that AI clients can read on their own |
.pdf | Text-based PDFs. Pages without a text layer, such as scans, have no searchable text. Images and drawings aren't included. Encrypted or password-protected PDFs can't be processed |
Files in any other format are refused before they upload.
Limits
| Limit | Value |
|---|---|
| File size | 25 MB per file |
| Files per upload | 10 at a time |
| PDF length | 500 pages |
| PDF processing time | 60 seconds per file |
| Extracted text | 8 MB per file |
| Knowledge storage | 1 GB per organization, for original files and their extracted text |
| Collection name | 100 characters |
| Collection description | 500 characters |
| Folder name | 100 characters |
| Folders | One level inside a Collection |
File statuses
| Status | Meaning |
|---|---|
| Queued | Uploaded and waiting to be processed |
| Processing | Text is being extracted |
| Indexing | Text is extracted and opens from its source page, but isn't searchable yet |
| Ready | Searchable by AI clients. A PDF with some image-only pages also shows how many pages have text |
| No text | No extractable text was found, so the file isn't searchable. You can still download it |
| Failed | Processing stopped. The reason appears under the status |
While a replacement is on its way, the row keeps the current version's status and adds the new version's, for example "New version processing".
Failure reasons
| Reason | What to do |
|---|---|
| Unsupported file type | Upload a text, Markdown, or PDF file |
| File contents don't match its type | Check that the file's contents match its extension, then replace it |
| Text isn't valid UTF-8 | Save the file as UTF-8 and replace it |
| PDF is password-protected | Remove the password or encryption and replace it |
| PDF has more than 500 pages | Split the PDF and upload the parts |
| PDF couldn't be read | Re-export or repair the PDF and replace it |
| Extracted text is over 8 MB | Split the document and upload the parts |
| Processing timed out | Simplify or split the PDF and replace it |
| File is too complex to process | Simplify or split the PDF and replace it |
| File is over 25 MB | Upload a smaller file |
| File size changed during upload, File changed during processing | Upload the file again |
| Knowledge storage is full | Delete files you no longer need, then replace it |
| Original file is missing, Storage was unavailable, Processing failed | Temporary. Click Retry processing in the file's row menu |