FAQ
Before you drop your first file.
Accuracy, pricing, languages, AI summaries, privacy, links, documents - everything people ask.
How accurate is the transcription?+−
Up to 99% accuracy on clear audio. Real-world recordings typically land at 95-98%: accents, crosstalk, and background noise lower that; a good mic raises it. Every word is timestamped so the last percent is fast to find and fix.
How does pricing work?+−
Subscribe monthly for the best rates - 30 hours for $10, 200 for $50, or 450 for $100. A file costs exactly its length: a 90-minute interview uses 1.5 hours. Plan hours reset each billing cycle; if you run out mid-cycle, top-up packs (5h for $3, 20h for $8) never expire and are spent only after your plan hours. Files can be up to 10 hours each, and failed transcriptions refund automatically.
What do I get for free?+−
3 files a day, up to 30 minutes each, in every language and every export format (with a small watermark line), plus AI summaries and questions. No card required. Paid hours kick in only for longer files or bigger days - they remove the watermark and add speaker labels.
Do transcripts come with AI summaries?+−
Yes - one click on any completed transcript generates an overview, key points, and action items, written in the transcript's own language. Summaries are included free on every plan, and if you share a transcript, the link shows its summary too. Summaries and transcript questions share your daily AI allowance (see the next answer).
Can I ask my transcript questions?+−
Yes - Ask this transcript answers in the language you ask in, and every claim cites the exact moments it came from: click a timestamp chip and the player jumps there. To keep the AI features free for everyone, questions and summaries share a daily allowance - 25 AI requests a day on the free plan, 300 on paid plans, resetting at midnight UTC. The allowance applies to every transcript, including ones transcribed on your device.
Which formats can I upload?+−
MP3, WAV, M4A, FLAC, OGG for audio; MP4, MOV, WEBM, MKV for video. We pull the audio track from video automatically - no conversion needed on your side.
Can I transcribe a YouTube or Instagram link?+−
Yes - paste a link from YouTube, Instagram, Facebook, X, TikTok, and 100+ other sites, or a direct video/audio file URL, into the Media downloader. You see the title, thumbnail, and length before confirming, and the transcript costs the same as uploading the file yourself. Public links only - private posts and live streams can't be fetched.
Can it read PDFs, Word files, and images?+−
Yes - drop a PDF, DOCX, image, or text file into the uploader. Digital documents extract instantly; photos and scanned pages (up to 20) go through AI OCR, with tables and headings preserved. Each document uses a free file slot or one minute of your balance and exports as TXT, DOCX, or PDF.
Can I record straight from my microphone?+−
Yes - the voice recorder records in your browser with pause and resume and a live level meter. Nothing leaves your device until you hit Transcribe, and the recording is priced like any uploaded file.
Is there a Chrome extension?+−
Yes - Flashscribe for Chrome puts your account one click from any tab: transcribe the video you're watching without sitting through it, record the call you're in (the tab's audio plus your microphone - nothing joins the meeting), or watch a channel for new episodes. Tick Transcribe on this computer and the recording never leaves your machine - only the transcript is saved. It uses the session you already have on the site; there is no separate login.
Can I transcribe without uploading my audio?+−
Yes - on computers with a capable GPU, the uploader offers Transcribe on this device: Whisper runs inside your browser, and your audio never leaves it - only the finished transcript is saved to your account. It's free and doesn't use your daily free files; plan length caps still apply (30 minutes per file on the free plan, 10 hours on paid). The model downloads once (~half a GB), and you keep the tab open while it works. AI summaries and questions work on these transcripts too, under the same daily AI allowance.
Can Flashscribe watch a podcast or a channel for me?+−
Yes - Sources. Paste a podcast, a YouTube channel, a blog, or any page and Flashscribe checks it every 30 minutes. New episodes and posts are scored 0-100 against your topics with a one-line reason, and articles get three to five facts, each backed by a quote that was verified to be in the text. Nothing is transcribed until you approve it - from the Sources page, or straight from the digest email (daily on paid plans, weekly on Free). The free plan watches 2 sources, paid plans 25.
Does watching a source use my hours?+−
No. Watching, scoring and fact-finding are free - they share your daily AI allowance (about ten items per AI request). Hours are spent only when you approve an episode or video, which is priced like any link, by its length. Saving an article costs nothing. Paid plans can turn on auto-approve for items scoring 85 or more, capped at 3 a day so it never runs away with your hours.
Which languages are supported?+−
90+ languages, auto-detected - including English, Spanish, French, German, Russian, Mandarin, Arabic. Mixed-language recordings work; each segment is tagged with its detected language.
Does it tell speakers apart?+−
On paid transcripts, yes - any file paid from plan hours or a top-up pack. Every turn is labelled Speaker A, B, C… - click a label to give it a name, and the names carry into every export and share link. You also see how long each person spoke, and can read or play back just one voice. Speaker labels cover the major languages; a solo recording simply shows none.
How fast is it really?+−
About a minute per hour of audio for an uploaded file. Links take a little longer, because the media is fetched first. Free-plan files queue behind paid ones at busy times but still typically finish within a few minutes.
Is there an API?+−
Yes - send a file, get a timestamped transcript back, using the same hours as your account. Create a key in Settings → API keys and see the API docs for curl examples. API usage always draws on paid hours - the free daily allowance applies only in the app.
Is my audio private?+−
Files are encrypted in transit and at rest, processed automatically, and never used to train models. Delete a transcript and the source file goes with it, immediately. For recordings that must not leave your computer at all, use Transcribe on this device - the audio is never uploaded.
Can I cancel anytime?+−
Yes - one click on the Plans page, no email required. Your plan runs to the end of the paid cycle and simply doesn't renew - and if you change your mind before then, resuming is one click too. Top-up credits you've bought stay on your account forever, and your transcripts stay downloadable on the free plan.
Still wondering?
The fastest answer is a transcript. About a minute, no card.