Turn every lecture, seminar, interview, or field recording into searchable text — then chat with Claude about any moment. Built for professors, researchers, grad students, and scientists.
What you get
Upload once. Skim in minutes instead of re-listening. Jump to any moment, export anywhere, ask AI the questions you'd ask a research assistant.
Every recording becomes a structured transcript with timestamps and chapter markers. Scan a 90-minute lecture in under five.
"What did I say about gradient descent?" → instant answer with a timestamp you can jump to. Use it like a research assistant who read the whole thing.
PDF, Word, Markdown, subtitles. Paste straight into Notion, LaTeX, Obsidian, or your course-materials folder.
Your audio is yours. We run open-source Whisper on private infrastructure — never sent to OpenAI or Google. Delete any file with one click.
Use cases
The job-to-be-done changes, but the mechanics don't: record or upload, skim, ask, cite.
Typical workflow
No setup, no plugins to install, no command line.
.mp3, .m4a, .wav, .mp4 — drag and drop, or paste a Zoom / YouTube URL. No file-size limit for paid tiers.
Speaker labels, per-word timestamps, chapter-style structure. A 60-minute lecture lands in roughly a minute.
Ask Claude about any moment, export to PDF / Word / Markdown, or share a read-only link with a colleague or student.
New · clipping for lectures
Drop the recording. Get the lecture transcript and a handful of 30–90 second clips back — captioned, cropped to vertical or square, and titled by the strongest line in each segment. Useful for asynchronous courses, conference shares, and short-form dissemination of research talks.
One full lecture, 4–6 short clips that map to the key concepts. Drop them into your LMS or share with students who missed the session.
Recorded a panel discussion or a Q&A? Split-screen layouts when 2+ speakers are on, single-speaker auto-zoom on the active one when only one is talking.
Word-level burned-in captions matched to your transcript. Edit any line in the transcript view and the caption updates with the clip — important for technical terms.
9:16 for Reels / Shorts, 1:1 for LinkedIn or X, 4:5 for Instagram feed, 16:9 for the course portal. Render once, share everywhere.
Pricing
Start free, scale up when your semester needs it. Credits never expire. Write to us with a .edu address for an education discount — we read every email.
1 hour / month · no credit card
Up to 100 hours a month
Volume + educational discount
Educational discounts available for verified faculty and .edu email addresses. Write to contact@neugence.ai — include your role and your institution, we'll reply the same day.
Frequently asked
Yes. We chunk long recordings automatically — there's no hard length limit. A 3-hour lecture transcribes in roughly 3–5 minutes on our Pro tier.
Yes. Paste a YouTube, Vimeo, or direct .mp3/.m4a/.mp4 URL and we pull the audio for you — no manual download step.
Yes. We run open-source Whisper on private infrastructure — your audio is never sent to OpenAI, Google, or any third-party transcription service. Delete any file at any time.
Every line of the transcript is anchored to a word-level timestamp. You can quote the speaker verbatim and link to the exact second in the recording.
Yes. Whisper handles 90+ languages out of the box, from Mandarin and Arabic to Hindi, Spanish, and Swahili. Mixed-language lectures work too — we auto-detect switches.
Upload it now — the first hour is on us, no sign-up, no credit card. Pick up right where you left off next semester.
Upload a lecture →