An upload that fails is almost always one of four causes. Find your symptom, apply the fix, then retry. Most uploads succeed on the second attempt once the cause is cleared.
| Symptom | Likely cause | Fix |
|---|---|---|
| “Upload failed” or a 500 error | Dropped connection or an oversized file | Retry on a stable connection; for big video, upload audio-only |
| File rejected at selection | Format not supported | Convert to MP3 (audio) or MP4 (video); see the format list below |
| Upload succeeds, then “no usable audio or video” | The file’s container is valid but the stream inside is empty or damaged | Re-export from the source app, or convert to MP3/MP4, then re-upload |
| M4A that won’t process | Codec variant Speak AI can’t read | Convert the M4A to MP3 and upload that |
Check the format first
Speak AI accepts MP3, WAV, OGG, WEBM, M4A, FLAC, M4P, AAC, MPEG, and AMR audio; MP4, WMV, AVI, M4V, MOV, and FLV video; TXT, DOCX, and PDF documents; and PNG or JPEG images (read with OCR). Anything else: convert to MP3 or MP4 with a free tool like HandBrake, then upload. Full details: File formats.
“No usable audio or video content”
The file looks fine, right extension and normal size, but the audio or video stream inside is missing, empty, or damaged. This usually comes from an interrupted export or a screen recorder that saved video with no audio track. Re-export from the original app; if you only have this copy, convert it to MP3/MP4, which rebuilds the container.
Upload works, but processing fails
That’s a different problem with its own causes. See Transcription errors.
Still stuck?
Send the file name, its rough size and duration, and what you saw on screen to success@speakai.co, or book a free consult and we’ll look at it with you.
Related: Uploads · File formats · Duration limits · Transcription errors