Upload Audio or Video to Transcribe Privately

How File Transcription Works

Choose an audio file or DRM-free video file, then Basil processes the media directly in your browser. Audio is decoded locally, video audio is extracted locally when your browser supports it, and multilingual Whisper transcription runs without a server upload.

After the transcript is complete, Basil uses the same local AI summary mechanics as the live web app. If your browser does not support WebGPU, transcription can still work, but the AI summary may be unavailable.

Private by Design

Your selected file is read by browser APIs on your device. Basil does not upload the media to a cloud server, does not keep a copy, and does not include the original media in the exported ZIP. When processing is finished, the page releases its reference to the file.

This browser tool is best for short meetings, interviews, lectures, and voice memos. For longer recordings, all-day capture, Apple Notes export, and more reliable media imports, use the native Basil AI app for iPhone, iPad, or Mac.

Browser Limits

• File size: v1 is limited to 100 MB to avoid browser memory failures.
• Video files: DRM-protected video cannot be processed. Some video containers may fail if the browser or local ffmpeg runtime cannot decode them.
• Performance: Whisper transcription and local summaries use your device. Older laptops may take several minutes for longer files.
• Best browsers: Chrome and Edge provide the most reliable model and WebGPU support.

Free Audio Transcription Tool for Private Files

Basil's upload transcriber is built for people who need a free audio transcription tool without sending recordings to a cloud transcription service. Upload a meeting recording, interview, lecture, consultation, podcast clip, field note, or voice memo, and the page converts speech to text locally in your browser.

Most online transcription tools require you to upload private audio before they can create a transcript. That is convenient, but it creates a copy of your file on someone else's infrastructure. Basil takes the opposite approach: the browser reads the file, the transcription model runs on your machine, and the exported transcript is yours to keep.

If you need live speech-to-text instead of file upload, use the private live meeting transcriber. If you want the full native workflow with longer recordings and Apple Notes export, see the private meeting transcription app.

Video to Text Without Cloud Uploads

This page can also work as a private video transcription tool for common DRM-free video files. When possible, Basil extracts the audio track locally, discards the media reference after processing, and turns the spoken content into a readable transcript and summary.

That makes it useful for recorded Zoom calls, exported webinar clips, training videos, customer interviews, research sessions, classroom recordings, product demos, and internal updates. Because video containers vary, some files may require the native Basil app or a different export format, but the goal stays the same: process sensitive recordings without uploading them to a transcription server.

For teams comparing privacy risks across AI meeting tools, our guide to AI meeting notetakers explains why upload-based cloud transcription can create unnecessary retention and compliance exposure.

Who Uses Private File Transcription?

A private upload transcriber is most valuable when the recording contains information that should not become another vendor's stored asset. Basil is designed for professionals who need fast transcription but still care about data ownership, confidentiality, and control.

• Legal teams: Transcribe client calls, deposition prep, interviews, and case strategy recordings while reducing attorney-client privilege risk.
• Healthcare and wellness professionals: Review consultation notes, intake recordings, and training materials without routing patient or client audio through a cloud workflow.
• Executives and founders: Convert board discussions, investor calls, hiring conversations, and strategic planning sessions into private meeting notes.
• Researchers and journalists: Turn interviews and source conversations into searchable text while keeping raw recordings local.
• Students and educators: Summarize lectures, seminars, study sessions, and classroom recordings without creating unnecessary online copies.
• Creators and consultants: Repurpose recorded conversations, coaching calls, and podcast drafts into summaries, outlines, and action items.

Private Transcription vs Cloud Transcription

Cloud transcription services can be helpful, but they usually require uploading audio or video to remote infrastructure. That means your file may be stored, processed by subprocessors, retained for debugging, included in audit logs, or governed by policy changes you do not control.

For deeper context, read why meeting transcripts should never touch the cloud and the hidden costs of cloud AI data retention.

What You Can Transcribe

The upload tool is intended for common, non-DRM media files that your browser can read. Browser support varies by operating system and file container, so a file that works in one browser may fail in another. For best results, start with short files and use Chrome or Edge on a desktop or laptop.

• Audio files: Voice memos, meeting recordings, interview audio, podcast drafts, lecture audio, and consultation recordings.
• Video files: DRM-free MP4, MOV, WebM, and similar exports where the audio track can be decoded or extracted locally.
• Meeting exports: Recordings from conferencing tools, screen recorders, phone recorders, and camera apps when you have permission to process the file.
• Languages: Multilingual Whisper transcription is used for uploaded files, making the tool useful beyond English-only meeting notes.

Always follow local recording consent laws and workplace policies before transcribing a conversation. For a practical overview, read are meeting bots legal? consent laws for AI notetakers.

How to Get Better Transcripts

Local AI transcription quality depends on the recording. A clean audio track gives the browser model more signal and less noise, which improves the transcript and the final AI summary.

• Use recordings with clear speech and minimal background noise.
• Export audio when possible instead of uploading a very large video file.
• Keep speakers close to the microphone and avoid overlapping speech.
• Split very long recordings into smaller files for faster browser processing.
• Review the transcript before using it as a legal, medical, or compliance record.

For recurring meetings, the native Basil app provides a more reliable capture workflow because it is built for long on-device recording, transcription, summaries, and exports from the start.

Frequently Asked Questions

Is this upload transcription tool really free?

Yes. The browser upload tool is free to use and does not require an account. Your browser downloads the local AI models it needs, then runs transcription and summarization on your device when supported.

Does Basil upload my audio or video file?

No. The upload tool is designed so your selected file is processed by browser APIs and local models. Basil does not receive the media file on a server. The exported ZIP contains the transcript and summary, not the original media.

Can I transcribe video files?

Yes, for many DRM-free video files. Basil first tries browser-native decoding and can fall back to local ffmpeg.wasm extraction. If a video container is unsupported or too large for browser memory, export the audio separately or use the native Basil app.

Why is there a 100 MB limit?

Browser-based AI transcription uses local memory and local compute. The 100 MB limit keeps the tool practical for common laptops and reduces the chance of a browser crash during decoding, extraction, or transcription.

Does it work offline?

The page needs an internet connection to load the site and AI model files. After the models are cached by the browser, some repeat workflows may work with fewer network requests, but the most reliable fully offline experience is the native Basil app.

Is this a HIPAA or GDPR transcription tool?

Basil's local-processing architecture helps reduce cloud data exposure because media does not upload to Basil servers. Compliance still depends on your organization, consent workflow, device security, and record handling policies. Learn more in our HIPAA compliant transcription guide and GDPR compliant meeting notes guide.