Transcribe a long NVIDIA app recording and keep only what you said.
A long NVIDIA app (ShadowPlay) recording is an MP4 you upload here, up to 10 hours or 5 GB, and a 4-hour session comes back as text in about 8 minutes. The speech is transcribed and the game audio is not; you get the first 20 minutes free, and two 4-hour sessions fit the $6 pack of 500 minutes.
Gameplay to script
Four hours of play hide maybe forty minutes worth quoting. The transcript shows you where they are.
Find the capture
Recordings from the NVIDIA app, and from ShadowPlay in GeForce Experience before it, are usually MP4 files under your Videos folder, often in a subfolder named after the game. The overlay's capture settings show the exact path.
Commentary, not the soundtrack
Music, gunfire and crowd noise are not transcribed; the speech is. If the game is loud and your microphone is quiet, the text holds what could be heard, so a capture with your mic recorded clearly gives the cleanest result.
Chapters per match
The chapters split a long session where the talk changes, from a ranked match to a build discussion to a viewer question, each with a title and timestamp you can turn into a highlight list.
Lines for the edit
Key quotes come out with their times, and the SRT and VTT files caption the edited upload. The upload page takes the MP4 as the NVIDIA app saved it.
Long sessions elsewhere
High-bitrate gameplay files are huge, and most text tools give up long before the session ends.
✗ The usual way
You rewatch the session at double speed and write down timestamps for the moments you remember.
- Captures many gigabytes large
- The OpenAI Whisper API caps uploads at 25 MB
- Rewatching four hours to find one call-out
- Game noise mistaken for words by weaker tools
- Subtitles typed line by line
- No list of what you promised viewers
✓ Here
Upload once, or send the audio when the video is too big, and get the session as text with its structure. Your voice is the part that becomes text, so the clearer the mic, the cleaner the transcript.
- Text about 8 minutes after a 4-hour upload
- Commentary transcribed, effects skipped
- Chapters per match or segment
- Timestamped quotes to guide your own cuts
- Subtitle files in SRT and VTT
- Two long sessions on one $6 pack
Sample output
Commentary over a long gameplay capture.
Written as an example of the format; not from a real stream, game or player.
Export
One transcript. Five clean formats.
Every pack exports all five — no format is held back for a higher tier.
Plain text
De-ummed paragraphs. Ready to paste.
SRT captions
Timed to the speech. Every video editor reads this.
WebVTT
HTML5 player + YouTube uploads.
Show notes
Formatted with chapters and pull-quotes.
Machine-readable
Timestamped text for your own tools.
Pricing
Honest pricing, no surprises.
A pack from $6 (500 minutes) opens the full transcript. After that, no subscription and no auto-renew — every tier is a one-time minute pack, and credits never expire. Buy once, use the minutes whenever you need them.
Casual pack
$6/500 min
8 hours of audio — $0.72 an hour. One-time — credits never expire.
- 500 minutes of audio
- All export formats
- Unlimited AI chat per transcript
- Executive summary & key quote
Best value
$12/2,000 min
33 hours of audio — $0.36 an hour. Credits never expire.
- 2,000 minutes of audio
- Priority queue
- All export formats
- Unlimited AI chat per transcript
Bulk pack
$24/5,000 min
83 hours of audio — $0.29 an hour. Credits never expire.
- 5,000 minutes of audio
- Priority queue
- API access for batch pipelines
- All export formats
FAQ
NVIDIA app recordings, answered.
How long does a 4-hour capture take?
Plan on about 8 minutes; processing runs at roughly 2 minutes for each audio hour. The longest public run at /proof/long-files is 9.31 hours in 1,033 seconds.
What does a week of sessions cost?
New visitors get the first 20 minutes free, and a pack transcribes the rest of a longer file from where the preview ended. A 4-hour session uses 240 minutes, so two fit the $6 pack of 500 minutes. Five 4-hour sessions are 1,200 minutes, a job for the $12 pack of 2,000 minutes.
My capture is far bigger than 5 GB. What do I do?
Gameplay bitrates are high: at 50 Mbps one hour is 50 x 3,600 = 180,000 megabits, 22.5 GB. Extract the audio and convert it to MP3 first; at 160 kbps an hour is 72 MB. Before the first pack uploads stop at 2 GB, and larger files start once a pack is on the account.
I streamed the session too. Can I use the VOD?
Yes. Paste the Twitch, Kick or YouTube link and nothing is uploaded. VODs longer than 6 hours start once a pack is on the account, and any single file stays within 10 hours, so a longer marathon goes in as two parts.
Can it handle commentary in Korean or Japanese?
Yes. The language is detected automatically among 99, Korean and Japanese included, and a session that swaps languages is read on its main one. Commentary that mixes English gaming terms into Korean is handled the same way.
Which exports are there, and is the footage kept private?
Sessions export as TXT, DOCX, SRT, VTT and JSON. Recordings are transcribed on our own servers and are never used for training. WhipScribe has no link to the NVIDIA app; it reads the file you send.
Related
More gameplay and stream captures
Upload the session. Find the moments worth clipping.
Transcribe an NVIDIA app recordingOperated by Neugence Technology Pvt. Ltd. · contact@whipscribe.com · Security · Privacy · Terms