Browser-guided audio-to-text conversion with EU-only processing and a multi-layer deletion guarantee.
Anonymous Direct-to-Cloud Upload. Five-Layer Deletion Guarantee.
Turn long-form interviews into searchable text without handing a source's voice — and everything they said — to an unknown server. Built for qualitative research, oral history, and investigative work.
Produce accurate, word-for-word records of meetings, hearings, and depositions where confidentiality isn't a preference — it's an obligation. No account, no persistent copy on a third-party server.
Transcribe fieldwork interviews and focus groups for your thesis or study, knowing your participants' recordings never take up permanent residence anywhere outside your own device.
Watch this short demonstration to see exactly how a recording moves from upload to a speaker-labeled, exportable transcript — while the audio itself stays on EU-only infrastructure and is wiped the moment processing finishes.
Transcription costs just 5 credits per minute of processed audio — drawn from the same universal credit balance you use across every tool on this platform.
Credit Packages & PricingFree credits for testing are available daily.
Your browser requests a time-limited, write-only Signed URL and uploads the recording directly to EU-based cloud storage. Your original filename is discarded in favour of a randomly generated identifier before upload.
Our server hands the job to an isolated processing worker running on Google Cloud in Frankfurt. The worker reads the recording from EU-based storage and transcribes it with AI services strictly pinned to EU infrastructure via hardcoded regional settings. Your audio is never used to train any AI model.
The isolated processing worker writes the finished transcript straight into secure EU cloud storage, and your browser retrieves it from there through a short-lived link. The transcript never passes through our server — not even through its memory — and our server has no permission to read it.
Soft Delete and Object Versioning are disabled at the storage level, so a delete command is final. On top of that, an immediate deletion of the recording by the processing worker as soon as the job ends, an automatic post-download deletion request, an hourly sweep for orphaned files, and a cloud-provider lifecycle rule — whose practical outer bound is close to 24 hours — make sure nothing is ever left behind.
Your transcript is assembled into a .txt, .docx, or .pdf file directly inside your browser. The exported document never returns to our server — it goes straight from your screen to your device.
I do long-form interviews for my own research — sometimes ninety minutes at a stretch, sometimes with people who trusted me with things they wouldn't have said in front of a microphone if they'd stopped to think about where the recording might end up. Typing up hours of audio by hand simply wasn't sustainable anymore, so I went looking for a transcription tool that could do it for me.
No. Your recording is uploaded directly to EU-based cloud storage via a temporary Signed URL, and our own server never receives the audio. Once your transcript is downloaded, five independent deletion mechanisms permanently erase the recording, backed by a cloud-provider lifecycle rule as the ultimate failsafe, whose practical outer bound is close to 24 hours. You can also read our detailed Privacy and Security Architecture document for a code-level deep dive.
No. Your recording and transcript travel only between your browser, EU-based cloud storage and an isolated processing worker. Our application server never receives them and is not granted permission to read or list them; it only exchanges job IDs and content-free status messages, and all storage access is audit-logged. The identity and permission design is documented in our Privacy and Security Architecture document.
No. We process your audio using AI API calls, restricted strictly to EU-only infrastructure. We have contractual guarantees that your submitted data is never used to train any artificial intelligence models — and any temporary storage during processing is erased by our five-layer deletion guarantee.
Our AI service provider enforces a small number of non-configurable safety filters, for example around certain sensitive personal data spoken aloud. If a recording is refused, you are not charged for it, any deducted credits are automatically refunded, and the uploaded audio is immediately and permanently deleted. Ordinary conversations — including heated arguments or sensitive topics — are not affected.
The tool accepts MP3, M4A, OGG, and WebM audio files. You choose the spoken language before processing so the AI transcribes rather than translates, and the finished transcript can be exported as a .txt, .docx, or .pdf file, generated entirely in your browser.
No registration, email, or password is required. Transcription costs 5 credits per minute of audio from your universal credit balance, and we provide free credits daily for testing, so you can try it before paying anything.
