How to Transcribe a Video File

To transcribe a video file, upload the recording, choose its audio language and whether you need speaker labels, then review the resulting text against the recording. You do not need to publish the video on YouTube or extract its audio first. EasyTranscriber accepts supported video files directly and transcribes speech from their first audio track.

Start with video to text for an MP4, MOV, or another video recording. If you already have an MP3, M4A, or other audio-only file, use audio to text. Both use the signed-in file-upload workflow described below.

File-upload guidance checked on 8 November 2026. Upload pricing and limits differ from the YouTube-link workflow.

What should you check before uploading?

Play the local file first. Confirm that you can hear the intended speaker at the beginning, in the middle, and near the end. A video can look complete while containing silence, a separate commentary track, or only background music.

Check these requirements:

• Account: sign in to upload audio or video.
• File size: no more than 500 MB per file.
• Duration: no more than 5 hours of audio per file.
• Content: the file must contain an audio track with speech to transcribe.
• Permission: only upload recordings you are allowed to process. Consider other speakers' consent and any workplace, course, or research restrictions.

The result is a speech transcript, not a description of the pictures. Slides, on-screen code, handwritten equations, and silent demonstrations are not transcribed from the video image. Upload transcripts have no timestamps, so this is not a timed subtitle-export workflow.

Which video and audio formats can you upload?

The file-upload form accepts the following formats. The file's actual type must match its extension; changing a filename from .avi to .mp4 does not convert it.

A supported container does not guarantee that every codec or damaged recording inside it can be processed. For example, an MP4 is a container: the encoded audio within it still needs to be readable. If processing fails, see the troubleshooting steps below rather than repeatedly uploading the unchanged file.

How do you transcribe the file step by step?

1. Open the file-upload form

Sign in and open the dashboard. Select Upload Audio or Video, then Choose an audio or video file. Select your local recording and check the displayed filename and size before continuing.

This is a file picker, not a field for a video-sharing link. For a YouTube URL, use the separate YouTube mode and the YouTube transcription guide.

2. Choose the audio language

Leave Audio language on Auto-detect, or select the spoken language from the available options. Choose the language being spoken, not the language you eventually want to write your notes in.

The optional Nova-3 Multilingual setting uses automatic language detection. It does not require additional credits. A language setting cannot restore speech that was not recorded or make an inaudible passage clear.

3. Decide whether you need speaker labels

Enable Speaker detection if you want the transcript separated with labels such as Speaker 1 and Speaker 2. This costs twice the standard transcription credits.

These are labels, not verified names. Listen to the recording before replacing a numbered label with a person's name, especially where speakers interrupt one another. For a single-person presentation, you may not need this option.

4. Select Transcribe file

Submit the recording with Transcribe file. EasyTranscriber processes the first audio track and returns the transcript. You do not need to create an audio-only copy of a supported video just to start.

Keep your original recording for review. If the text contains the wrong language, only music, or an unexpected speaker, check which audio track is first rather than assuming the model heard the same track your video player selected.

5. Review and save the text

Read the transcript while replaying the recording. Pay particular attention to names, technical terms, numbers, dates, negation, and speaker attribution. Check every passage you plan to quote; a plausible sentence can still misrepresent the speaker.

Download options support TXT and Markdown, according to your export preference. They do not add timestamps or turn the upload transcript into synchronized captions. Keep a separate edited copy if you remove repetitions or add observations from the video image.

For a recorded class, the next step is organizing and checking the content, not just shortening it. The lecture-to-study-notes guide explains how to add missing visual context and make useful review questions.

How many credits does a video file cost?

File uploads cost 1 credit per started hour of audio, or 2 credits per started hour with speaker detection. The service measures the decoded first-track audio duration and reconciles the charge against the processed duration. File size and video resolution do not determine the credit cost.

A free account starts with 5 credits and refills by 1 credit per day up to a balance of 5. Check your balance before choosing speaker detection for a long recording.

Rounding applies to each submitted file. Splitting 61 minutes into two files of 30 and 31 minutes still costs 2 standard credits in total. Compressing a video without shortening its audio can solve a size problem, but does not reduce its transcription duration or credit cost.

Are summaries and chat included?

For an uploaded file, request the first AI summary at no extra credit after transcription. It does not start automatically. Regenerating the summary costs 1 additional credit. Review a summary against the transcript and original recording before relying on it.

Chat is optional: signed in, you can ask questions about the transcript (1 credit per 2 messages).

Why did the upload fail or produce the wrong text?

The file is over 500 MB

Export a smaller copy, or export the intended audio as MP3 or M4A and use audio to text. Preserve a clear speech track rather than reducing audio quality aggressively. Keep the original video if you will need to review slides or demonstrations later.

Size and duration are separate limits: a small file can still exceed 5 hours, and a short high-resolution video can exceed 500 MB.

The audio is longer than 5 hours

Trim the recording to the section you need, or split it into files that each meet both limits. Keep the parts clearly named and in order. Remember that each file is billed in started-hour increments; splitting is not necessarily cheaper.

The recording plays, but the transcript follows the wrong track

A video player may let you select among several audio tracks. EasyTranscriber processes only the first audio track. If that track contains commentary, another language, or music instead of the intended dialogue, export a new file with the desired speech as its only audio track and upload that copy.

The same issue can affect screen recordings with separate microphone and system-audio tracks. If you need both sources, use your editor to combine the required sound into one track, listen to that export, and then upload it.

The file has a supported extension, but cannot be decoded

Re-export from the application that created the recording, or use a media editor to export the desired audio as MP3, M4A, or WAV. Check playback of the exported file before trying again. Renaming the extension does not repair an unsupported codec, incomplete download, or damaged file.

Keep an eye on the size limit when exporting WAV: an uncompressed audio copy can be larger than a compressed one.

The video is silent or has no speech

Check the recording's microphone and system-audio capture settings. If the original file contains no audio track, there is nothing for speech recognition to transcribe. If it contains only music or silence, it cannot supply a spoken-word transcript.

Changing formats cannot recover sound that was never recorded. On-screen words would require a separate visual-text workflow; this upload feature does not perform OCR.

The words or speaker labels are inaccurate

Replay the affected passage and check the language setting and selected audio track. Background noise, overlapping speech, and unfamiliar names can make the result harder to use. Mark an uncertain word rather than silently inventing one, and verify numbered speaker labels before assigning names.

For important quotations, check the original audio yourself even if the rest of the transcript reads fluently. No format choice or speaker-label setting removes that review step.

What happens to the uploaded file?

We delete your uploaded file from our servers after processing. Deepgram, our speech-to-text provider, does not store the audio or transcript or use it to train models. We save the transcript to your account.

Uploading a file is not offline processing. Before sending confidential material, check your organization's requirements and whether you have permission to use an external transcription service. The privacy policy provides broader information about account data.

Which workflow should you use next?

Use video to text when the recording is still a video file, or audio to text when you already have its audio. You do not need to convert a supported video solely to upload it.

For source-specific instructions, see voice memo transcription, Zoom transcription, or Google Meet transcription. Those pages cover their recording workflows; this guide focuses on processing a local file you already have.