Speaker separation is conditional
Managed recognition can estimate speaker turns. The local engine has no diarization, and overlapping speech can confuse even a diarization model. Verify and rename any speaker labels yourself.
✳BE PRESENT. COME BACK TO THE DETAILS.
Interview Transcription With Timestamps
Stay with the conversation now and find the exact words later. Interview transcription turns an authorized recording into a reviewable transcript.
Sign in for a free trial
60 seconds free after signing in
Works with
YouTube
InstagramAudio & video files
YouTubeInstagramAudio & video filesFROM LINK TO LIGHTBULB
Interview transcription in three steps
A simple workflow that gets out of your way.
01
Bring your recording
Paste a video URL or upload a file that you own, that is authorized or publicly accessible.
02
Let the words appear
We read available captions or transcribe the spoken audio of authorized or publicly accessible recordings only.
03
Make something of it
Review, copy, or export text and subtitles from your authorized or publicly accessible recording.
MORE THAN A WALL OF TEXTThe whole thought.
The whole thought.
Ready to work with.
Readable words. Useful timestamps. A starting point for whatever comes next.
- Find the moment
Search the words and keep their timestamps.
- Take it anywhere
Copy text, or export TXT, SRT, and VTT.
- Make it yours
Review and edit the draft before you use it.
Treat the recording as the source of truth
Add a small vocabulary
Supply participant names, product names, or specialist terms separated by commas. For example: Acme, onboarding, retention cohort. Recognition hints improve context without guaranteeing exact wording.
Consent and retention
Obtain permission from everyone whose recording you process. Media is removed after processing; results expire after 24 hours or can be deleted immediately. This MVP is not a certified medical, legal, or regulated-data service.
Supported files: MP4, MOV, AVI, MKV, WebM, MP3, WAV, M4A, FLAC, OGG, AAC · 100 MB maximum. Free trial: 60 seconds total; credits: up to 60 minutes. Original-language transcription, with automatic language detection. Accuracy varies by language and recording.
THE BOUNDARIES, UP FRONT
What You Can Submit
- Interviews where every participant agreed to be recorded and processed.
- Speaker separation: managed recognition can estimate speaker turns (Speaker 1, Speaker 2). Labels are not identities; review them.
- Retention: the recording is deleted right after processing, and the transcript expires after 24 hours. Delete it sooner with Delete transcript on the result page.
- Interview recordings and transcripts are never used to train our models.
Text and subtitle files only — no media downloads. Read the acceptable use policy.
START SMALL. GO FURTHER.
Try it free. Top up when you need more.
No subscription to keep track of. Try one minute free, then add minutes when you need them.
| What’s included | Free trial To see the quality | Prepaid minutes For longer recordings |
|---|---|---|
| Allowance | One-time 60 seconds per user | Pay once for the minutes you use |
| Recording length | Up to 1 minute | Up to 60 minutes |
| Timestamps & exports | TXT, SRT, VTT | TXT, SRT, VTT |
| Sign-in | Required | Required |
| If a task fails | Allowance returned | No minutes used |
PAY ONCE · NO SUBSCRIPTION
Need more minutes? Top up once.
One credit covers one started minute of audio. Buy a pack once and use it whenever you like for 12 months.
100 minutesAbout 1.5 hours, or up to 100 short clips
$4.90one-time · 4.9¢ per minute
1,000 minutesAbout 16 hours of recordings
Best value for most$19.90one-time · 2.0¢ per minute
3,000 minutesAbout 50 hours of recordings
$39.90one-time · 1.3¢ per minute
- Recordings up to 60 minutes
- Timestamps, TXT, SRT and VTT
- Speaker labels when speech recognition runs
- Valid 12 months; a new top-up extends it
- Failed jobs never use minutes
Transcription output only. No media downloads. Prices in USD · Secure checkout with Stripe · Any applicable tax is shown at checkout before you pay. Sign in to buy. Manage wallet
THE LITTLE THINGS, EXPLAINED
A few good questions.
Clear answers before you press transcribe.
What kind of video links can I submit?+
Public, directly accessible links to videos you own, that are publicly accessible, or that you are authorized to process. We don't support links that require a login.
Do you keep or redistribute the original video?+
No. To transcribe a link or upload, we fetch the audio and process it temporarily. The media is queued for deletion when processing finishes; cleanup may take a few minutes, and we never offer it for download or redistribute it. You receive transcript text and subtitle files only.
Can I process private or login-protected videos?+
No. Private, login-protected, deleted, restricted, or otherwise inaccessible content is not supported.
Are you affiliated with Instagram or YouTube?+
No. textfromlink is an independent tool. Platform names are used only to describe compatible input types.
Does interview transcription automatically identify the people speaking?+
No. When the configured engine supports diarization, it can label estimated speaker turns as Speaker 1, Speaker 2, and so on. It does not know who those people are. Local Whisper output has no speaker labels.
What is included in the free allowance?+
Sign in for a one-time 1-minute trial. Each user receives a cumulative 60-second allowance that does not reset each day, on another device, or after account deletion. Failed jobs release unused reservations.
How much do longer transcripts cost?+
Credit packs start at $4.90 for 100 credits. One credit covers one started minute of a completed recording, up to 60 minutes per recording. Failed tasks never cost credits.
How accurate is the transcript?+
Accuracy depends on audio quality, accents, and overlapping voices. Copy the text or export TXT, SRT, or VTT with timestamps, then review names, numbers, and timing before publishing.