Frequently asked questions

Create a free account with your email address — we send a code, there is no password and no card. Your first transcription is free whatever its length; after that you get 5 minutes a month, or pick a plan. Upload a file on the dashboard, or create an API key under “API & connectors” to use it from code.

No. Everything happens in the browser: upload on the dashboard, download the files when they are ready. Developers can use the same service from code or an AI assistant instead.

Common audio and video formats — MP3, M4A, WAV, FLAC, OGG, MP4, MOV, WebM and more. Upload the file, or pass a public URL and we fetch it. YouTube video links work too: we import the video’s captions in seconds.

YouTube, yes: paste a video link and we import its caption track, usually in seconds, as the same TXT, SRT, VTT, JSON and PDF files. Captions carry no speaker labels, and a video without captions cannot be imported (download its audio and upload that instead). Each plan includes a number of YouTube videos a month, 2 on the free account. Spotify, SoundCloud, Apple Podcasts and Drive share links are web pages, not files: download the file and upload it.

Yes, at audiotranscribr.com/app: upload files, download transcripts, see your usage, change plan and create API keys. Sign in with your email; we send a one-time code, there is no password.

Up to 5 GB per file, audio or video, by upload or URL. Multi-hour recordings are fine — a 2.7-hour, 76 MB file is part of our test set and finishes in under a minute. Large videos have their audio track extracted first, which adds a few minutes. A job that is still processing after one hour is cancelled and not billed.

Yes, the dashboard works in your phone’s browser. On an iPhone or iPad, a large video picked from the Photos library can fail to upload: the phone repackages it first, and the browser can run out of memory. Save the video to the Files app (in Photos: Share, then Save to Files) and choose it from there, or upload from a computer. Keep the screen on and the browser open until the upload finishes: phones pause uploads in the background.

Every job produces five files: JSON (words, timestamps, speakers, confidence), SRT and VTT subtitles, plain text, and a formatted PDF transcript. Each transcript also gets a short summary in the language of the recording, shown on the dashboard and at the top of the text and PDF files. Download any or all of them.

Yes, on every plan, including Free. Speaker labels appear in the JSON, the text file, the VTT voice tags and the PDF.

Short clips come back in seconds. Longer files take longer, but you never wait on an open connection: create the job, then poll its status.

Leave the language out and it is detected automatically — 35 languages, including English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Polish, Turkish, Hindi, Japanese, Korean, Chinese, Indonesian and Vietnamese. Set it explicitly (es, pt-BR, ar, he, zh-TW …) to reach 60+ languages, including Arabic, Hebrew, Bengali, Tamil and Urdu. Speaker labels work in every language.

Bad input (a page instead of a media file, an unknown plan, invalid JSON) is rejected immediately with a 4xx and a machine-readable error code. If the media cannot be fetched or decoded, the job ends with status “failed” and an error field, and you are not billed for it. The API reference lists every code and which ones are safe to retry.

Plans include a monthly allowance of audio. On paid plans, extra minutes are billed per second at the plan's overage rate, so a busy month never stops your app. The free account gets one transcription free, then 5 minutes a month, and stops there. Change or cancel any time.

Uploaded media is deleted automatically after 7 days and transcripts after 30 days. You can delete a transcription and its files immediately with one API call.

Use “Recover your API key” on the pricing page. We email a secure link to the address you signed up with; keys themselves are never sent by email.

Yes. The API is also a remote MCP server: add one URL and your API key to Claude Code, Cursor, VS Code or any MCP client, and the agent gets tools to transcribe media, read transcripts and fetch the subtitle and PDF files. In Claude.ai or ChatGPT, add the server as a connector at https://api.audiotranscribr.com/mcp and sign in with your email, no key needed.

You don’t need one — it’s four REST endpoints and an API key header. The docs have copy-paste examples for curl, JavaScript, Python and PHP, plus an OpenAPI file for code generators and a public Postman workspace you can fork.