01
Bring the recording
Upload a file or import a link. Large files resume where they left off.
Transcription for thoughtful work
Turn recordings into useful knowledge.
Word-level transcripts, a focused editor, and AI tools that help you understand interviews, lectures, and conversations.
Word-level timing Speaker separation Private storage
Upload preview
One calm workflow
From raw audio to something you can act on.
02
Follow every word
Seek by word, name speakers, and edit while the audio stays in sync.
03
Shape the result
Export subtitles and documents, translate, or create cited notes.
Product preview
field-interview.m4a
42:18 · English · 2 speakers
Maya Chen
The useful part begins when the recording becomes searchable, editable, and easy to cite.
Daniel Okafor
Exactly. A transcript should feel like a working document, not a wall of text.
00:22
42:18
One source, several working views
The recording remains the reference point.
01 / Listen
Trace text back to sound
Word timing, speaker structure, and capability-checked multi-engine routing.
02 / Read
Review every subtitle cue
Shape SRT and WebVTT with visible reading-speed and line-length checks.
03 / Synthesize
Keep notes close to evidence
Summaries, chapters, actions, and conversation tied to transcript citations.
04 / Carry
Move across languages deliberately
Resumable translation with a pinned source revision and bilingual export.
Reviewed directories
Facts first. Drafts stay backstage.
Each directory is generated from the same reviewed catalog used by routes and sitemaps.
30 reviewed
Browse format guides Compare containers, MIME boundaries, source preparation, and limitations across every reviewed format guide.98 reviewed
Browse language guides Find writing-system and review guidance only for languages inside the executable deployment snapshot.12 reviewed
Browse subtitle workflows Open independent cue-preparation and QA workflows for every reviewed subtitle conversion path.Reviewed workflow library
Pick a path by job, not by keyword.
Every card leads to a finite, typed directory. Unreviewed drafts stay out of routes and sitemaps, while local tools expose their processing boundary before a file is selected.
16 reviewed
Compare transcription products
Eight source-checked comparisons and eight reversible switching guides.
Browse directory10 reviewed
Browse audience workflows
Capture, review, privacy, and handoff guidance for distinct real workflows.
Browse directory40 reviewed
Review translation pairs
Timed subtitle workflows with script, direction, line-break, and bilingual QA notes.
Browse directory8 reviewed
Open browser-local tools
Local subtitle, text, audio, and video utilities whose file bytes stay in the browser.
Browse directoryReviewed fixtures, not flags
Names, native names, and stable language tags.
These six language fixtures are frozen in the current transcription acceptance set. Availability still depends on tier, model, and requested features.
- en
- EnglishEnglish
- zh
- Chinese中文
- es
- SpanishEspañol
- ar
- Arabicالعربية
- hi
- Hindiहिन्दी
- ja
- Japanese日本語
Limits without fine-print theater
Know the boundary before the upload.
Start with daily free limits, then move to Pro for larger files, batch work, every reviewed tier, and priority scheduling under fair use.
- Monthly
- $15.00USD
- Yearly
- $96.00USD
Before you upload
Clear answers, including the edges.
What does the private trial include?
The anonymous trial accepts one supported audio or video file up to 100 MB and 10 minutes, then physically clips the processing media and returns only the first 90 seconds.
How long is trial data retained?
Unclaimed trial objects and their trial metadata expire after 24 hours. Saved account data follows the account export, trash, and deletion controls described in the privacy draft.
Are timestamps and speakers always exact?
Word timing and speaker separation depend on the selected model and recording. The interface keeps those capabilities explicit and expects review against the audio.
Does a language name guarantee every model supports it?
No. The current deployment capability snapshot decides which reviewed models can serve a requested language, tier, and feature set.
Bring one recording. Leave with inspectable words.
Upload one supported audio or video file. The private preview is bound to this browser and expires after 24 hours if it is not claimed.
Try a recording