How to Upload Audio and Video Files for AI Transcription

A podcast episode, recorded interview, meeting, or lecture often starts as a file on your device. You do not need to publish it on YouTube to turn it into a transcript. Transcript Guru can upload the recording and run AI speech recognition, or ASR, directly on its audio.

Upload a recording for ASR transcription

1. Sign in and open Transcribe in the dashboard.
2. Drop an audio or video file into the uploader, or choose it from your device.
3. Leave Language on Auto-detect, or select the language spoken in the recording.
4. Choose Review & transcribe to see the estimated credit cost, then select Transcribe now.
5. Open the job from Your uploads to read and review the finished transcript.

Supported formats include MP3, MP4, WAV, M4A, WebM, MOV, MKV, OGG, FLAC, and MPEG-TS. Basic supports files up to 500 MB and 2 hours. Pro supports files up to 2 GB and 4 hours. The uploader checks file size and type before transfer, then checks actual duration after upload.

Understand the credit estimate

ASR cost is based on recording length: one credit per 80 seconds, rounded up. A 40-minute recording is roughly 30 credits. The review screen uses local media metadata to estimate the cost. Some file formats do not expose a duration in the browser, so the estimate can be unavailable until the server probes the uploaded file. The server checks the exact cost before it queues the job.

Make sure your recording fits within your plan's size, duration, and available credit limits. You can see your balance in the review screen and manage your plan from Billing.

Transcribe several files in one batch

For a series of interviews or episodes, use Batch transcribe files on the same Transcribe page. Choose several recordings, confirm the shared language and estimated credits, and queue the batch. Each row shows upload progress, an error if that file could not be queued, or a link to its transcript. You can retry failed files without resubmitting the successful ones.

The number of files you can queue at once depends on your plan's open transcription slots. Other active jobs use those slots too. Keep the page open while files are uploading and being queued. Once a row says Queued, its transcription continues even if you leave the page.

Get a cleaner transcript

Clear audio makes the biggest difference. If possible, use the original recording rather than a compressed copy, keep voices close to the microphone, and reduce background music. Select the spoken language when you know it. After the job finishes, review proper names, product names, and technical terms before sharing or exporting the text.

A transcript makes a recording searchable and easier to reuse. You can pull quotes from an interview, write podcast show notes, or export captions such as SRT and VTT for video. For recordings already on YouTube, see our guide to transcribing a YouTube video.

Frequently Asked Questions