No file loaded
Ready
AI-Assisted Audio Editing

Wise voice tools for your podcast, videos, voice-over, audiobook, interviews or dubbing.

Online voice editor featuring AI restoration, transcript-based editing, show notes and social posts, voice cloning, classic studio tools, and Owliver—your AI assistant.

wav · mp3 · m4a · flac · aac · video
Hear the differenceA few examples
Before · echoey0:10
Fixed with Remove ReverbReal clips · no edits
Scroll
Transcript AI

Edit audio by editing text.

One click transcribes audio, detects languages, and separates speakers into an editable script.

  • 01Precision audio trimming.Jump anywhere by clicking a word, or strike out text to instantly clip the sound. Hunt filler words and delete them.
  • 02AI voice replacement.Fix typos without getting behind the mic. Re-voice creates seamless replacement audio in seconds.
  • 03Automatic speaker detection.Assign names to speakers once. hoot syncs those labels directly to your Episode Kit and final .txt export.
1 AI minute per minute of audio · Export .txt, .srt or .vtt
launch_interview.wavEN · auto
Episode Kit AI

You recorded it. hoot writes the rest.

Episode Kit reads your transcript and writes everything you need to publish and promote the episode: titles, show notes, chapters, social posts, shorts to clip, quotes, a newsletter, a blog post, tags and FAQs. One click, all from what was actually said.

ep42_nora.wav · 54:1213 sections ready

13sections from one transcript, written in one go

01Titles

PodcastYouTubeSpotifyPodcasts
  1. ANora Field: from a garage oven to 300 loaves a day
  2. BOne product, sold out daily: how a bakery starts small
  3. CThe fire on opening day, and what came after

02Summary

FormalCasualConciseEngaging

Nora Field baked her first loaves in a garage during a long winter. Four years later her bakery sells 300 a day. She tells Sam about the oven loan, the fire on opening day, and why the shop still closes on Mondays.

03Key points

  • Sell out every day before you add a second product.
  • The first loan paid for the oven, not the shop.
  • One closed day a week kept the team (and the starter) alive.

04Chapters

  1. Cold open
  2. Twelve loaves in a garage
  3. One product, sold out daily
  4. Fire on opening day
  5. Why Mondays stay closed
  6. Advice for new owners
Add as markersYouTubePodcast JSON

01Social posts

TikTokInstagramFacebookXLinkedIn
Post 01189 / 280

Her first 12 loaves came out of a garage oven. Now it’s 300 a day. 🍞 Nora Field on the oven loan, the fire on opening day and the Monday rule. New episode out now. #smallbusiness #sourdough

02Shorts to clip

  1. 1Why one product beats ten8:31–9:15 · 44sInsightful
  2. 2The oven fire on opening day24:05–24:52 · 47sFunny
  3. 3The week she almost sold the shop33:10–33:58 · 48sEmotional

03Quotes

“Sell out every day before you add a second thing to the menu.”
Nora Field

04Newsletter

SubjectThe garage oven that became a bakery

Hi there, this week Sam talks with Nora Field, who went from 12 loaves in a garage to 300 a day. She shares the one rule that kept her shop alive in year one…

01Search

Keywords
sourdough bakeryopening a bakerysmall business podcastbakery startup costsNora Field
Meta description 148 / 160

Nora Field turned a garage oven into a bakery that sells 300 loaves a day. Hear how she paid for the oven, got through a fire and kept Mondays free.

02Tags

sourdoughbakerysmall businessentrepreneurshipfood podcaststartup costsNora FieldSmall Shop Talk
8 tags · 109 / 500 characters (YouTube)

03Blog post

1,180 words · 6 min read · Markdown or HTML How a garage oven became a 300-loaf bakery

Nora Field never planned to open a shop. She planned to get through a long winter, and bread gave her hands something to do. Four years later, the queue starts before the door opens.

04FAQs

How did Nora Field pay for her bakery?
A small loan bought a deck oven. She signed a lease only after three months of selling out.
Why does the bakery close on Mondays?
One day off a week keeps the team rested and the starter on schedule.

05Mentions

  • PersonMarta Silvaher first hire
  • PlaceRiverside Marketthe first stall
  • ProductDeck oventhe loan
  • 01Every quote is in the audio.Quotes, shorts and mentions keep their time and speaker. Click one to hear it.
  • 02Chapters land on your waveform.Add them as markers, or copy them for YouTube or as Podcasting 2.0 JSON.
  • 03Shorts you can hear first.Play each moment, select it on the waveform, then trim and export the clip.
  • 04Your guests, by name.Name the speakers in the transcript once. The whole kit uses their names.
Needs a transcript · A full kit for a 1-hour episode is 13 AI minutes
Voice Clone AI

Your voice, any language.

Import text from Transcript or type it yourself. The character of the reference voice is retained and a credible clone is generated.

  • 01Speaker aware cloning.Detects multiple speakers and clones voice without any extra setup.
  • 02Translation control.Auto-translates scripts into editable previews for complete control.
  • 03Global reach made effortless.Produce episodes in up to 15 languages using your original voice, saving time and studio budget.
1 AI minute per 1,000 characters
Reference · the clip in your editor
0:06
Audio processing tools

Wise tools, specific tasks.

Clean up your track, shape your tone, and get creative with effects. Most tools run instantly in your browser and come with every Pass — the AI ones are clearly marked.

Clean up

Voice Cleaner

Light included · Pro AI

Isolates the voice from the rest — hiss, hum, fans, traffic, the room itself. The gaps go properly quiet.

Remove Reverb

Light included · Pro AI

Removes reverb and room tone to transform echoey recordings into clean, studio-dry vocals.

De-esser

Included

Attenuates harsh ‘S’ and ‘T’ sounds without dulling the rest of your vocals.

Auto-Cut

Included

Cuts the dead air between words but keeps a little padding, so the edit still breathes.

Gate

Included

Mutes everything below a threshold, so the gaps between phrases go truly silent.

Tone, dynamics & effects · all included, all in your browser

Parametric EQ

Shape the tone band by band.

Compressor

Even out the loud and the soft.

Saturator

Add warmth and analog grit.

Limiter

Catch the peaks before they clip.

Loudness

Hit the target level for any platform.

Reverb

Place the voice in a room.

Delay

Echoes, slap-back and space.

Speed

Faster or slower, same voice.

Got multiple or long audio files? No problem

Batch Processing

Edit one file and hoot records the effects chain. Open Batch Processing, add the other files, and click Apply Chain to All.

Multiple regions

Detect multiple regions in a single audio file. Edit them, normalize them, or export them in one go.

Markers & chapters

Tap M to flag a moment as you listen. Name and color the markers, then export each segment.

Owliver Beta

Name the issue. Owliver fixes it.

Describe the problem in plain words. Owliver picks the right tools, dials in the settings, and applies them to your audio.

  • 01Hands-free processing.Ask for cleaner, louder, or tighter audio. Owliver chains multiple tools in a single step.
  • 02Built-in tool guide.Click any panel name for a plain explanation of what the tool does and when to use it.
  • 03Bring your own model.Prefer another LLM? Plug in an Anthropic or OpenAI-compatible key from the cog.
Included in every Pass · AI tools he applies use AI minutes
Pricing

A whole studio for €4 a month.

Try the full editor for 3 exports. Then one small Pass keeps it open. AI tools use AI minutes — one minute of audio is one AI minute.

Free trial
€0

No card needed

  • The full editor + 17 free tools
  • 3 exports
  • AI tools not included
Most popular
Studio Pass
€4/month

€48 billed yearly · or €5 monthly

  • Full editor, unlimited exports
  • 150 AI minutes every month
  • Unused minutes roll over 1 month
Pro Pass
€12/month

For weekly shows · cancel anytime

  • Everything in Studio Pass
  • 600 AI minutes every month
  • Lowest price per AI minute
Need more AI minutes?Add a top-up to any Pass. Top-ups never expire while you have a Pass.
100 min€5400 min€151,000 min€30

Failed AI jobs give their minutes back · Cancel anytime · Prices include VAT

Your best take is one click away.

Pick a file, or drop one anywhere on this page.

wav · mp3 · m4a · flac · aac · video (extract audio)

Fix any recordingin your browser

Tap here to browse

wav · mp3 · m4a · flac
Time0:00.000/0:00.000
Drop your sample here, or click to browse files
Peak
-∞
LUFS -∞
100%
0:00.0
0:00.0
Audio Tools
Cleanup
Tone & Dynamics
Effects

About hoot

What is hoot?

hoot is a browser-based audio editor built for speed, simplicity, and privacy. Editing runs entirely in your browser — your audio stays on your device. The optional AI features (transcription, neural vocal isolation, dereverb, voice cloning and Episode Kit) are processed by online AI services and use AI minutes, which come with a Studio or Pro Pass; everything else is fully local.

Who built this?

hoot is built and maintained by Fábio Braz — independent designer, developer and audio enthusiast. It started as a personal tool for vocal editing and grew into a full-featured editor for podcasters, voice-over artists, and anyone who works with spoken-word audio.

What can it do?

Parametric EQ, compression, saturation, brick-wall limiting, de-essing, gating, dereverb, neural noise reduction, loudness mastering to broadcast and streaming standards, pitch-preserving time stretching, silence removal, region-based slicing, markers, batch processing, transcription, voice cloning, and an AI assistant that builds processing chains from a plain-language description. Open common audio formats, export to WAV or MP3 at standard bit depths and sample rates.

The approach

One screen. No projects-and-tracks bureaucracy. An editor that gets that you just want to fix a recording and move on.

Privacy Policy

Last updated: October 2026

The short version

hoot is an editor that runs inside your browser. Opening files, editing, playback, analysis, effects and export all happen on your device. Your audio leaves your browser only when you choose a feature that needs it — Transcript, Pro vocal isolation, Remove Reverb, Voice Clone or Episode Kit — and then only to the AI service that runs that feature, for the length of that one request.

Your audio

While you work, your audio lives in your browser's memory. hoot also keeps one automatic recovery snapshot in your browser's own local storage, so a crash or an accidental refresh doesn't cost you the session. That snapshot stays on your device and is cleared when you discard it, or once you export and close hoot.

The server-side features send a copy of your audio over an encrypted connection, process it and return the result: Transcript for speech-to-text, Pro vocal isolation and Remove Reverb for cleanup, and Voice Clone, which uses a short voice sample — the audio you loaded, or a reference clip you supply — together with the text you type, which is translated first if you pick another language. Episode Kit sends only the text of your transcript, not your audio, to an AI service that writes the notes. hoot does not keep these uploads, sell them, or use them to train models. The services that carry out the processing follow their own retention and logging policies and may log ordinary request and network data.

Project files

Project files (.hoot) are saved to and loaded from your own file system. They hold your audio together with your editing decisions — slices, markers, processor settings, transcript — in a single container, so opening one restores everything at once. They are never uploaded.

The assistant

To answer you, the assistant sends your messages, the recent chat history, the file name and measurements taken from your audio — things like duration, levels and tonal balance — to an AI model. Your audio file itself is never sent. If you connect your own AI provider instead, the key stays in your browser and those requests go directly to that provider, whose terms, retention and training policies then apply.

Your account

Server-side features need an account. When you sign in, we store your email address, your plan, and your AI-minute balance with a record of minutes added and used. Your display name and picture come from your sign-in provider. We keep them only in your sign-in cookie, not in our records. The cookie keeps you signed in.

Other services handle some of your data for us. Our payment processor, Stripe, handles payments. It keeps your email and payment records, and hoot never sees or stores your card details. If you send us feedback, we get your message and your email address. If you sign in with an email code, an email service sends that code to you.

To delete your account and its data, email us. We will delete your records, and we will ask Stripe to delete yours.

Analytics

hoot counts events, not people: page loads, demo plays, exports and AI passes, with no name, email, account ID, file name, audio or message content attached. As on any website, ordinary network information such as your IP address is handled in order to deliver the request.

Third parties

hoot loads some fonts, code and models from public networks, and relies on outside services for hosting, sign-in, payments and the AI features above. Those requests carry no audio beyond what the features you choose upload, but the providers may receive ordinary network data such as your IP address under their own privacy policies.

The services we use are: Cloudflare (hosting and servers), Google (sign-in and fonts), Stripe (payments), Resend (emails), AssemblyAI (Transcript), Modal (vocal isolation and Remove Reverb), Fish Audio (Voice Clone), Groq (the assistant and Voice Clone translation) and OpenAI (Episode Kit).

Contact

Questions about this policy: hey@hoot.audio.

Terms of Use

Last updated: October 2026

Acceptance

By using hoot, you agree to these terms. If you don't agree, please don't use the application.

What hoot is

hoot is a browser-based audio editor offered on a freemium model. The core editing features are free to use. Certain advanced and AI-assisted features are available under a paid plan. The app is provided "as is" without warranties of any kind, express or implied.

Your content

You retain all rights to any audio you edit with hoot. Core editing runs entirely in your browser, so that audio never leaves your device. The optional AI features (transcription, neural vocal isolation, dereverb, voice cloning and Episode Kit) send your audio to the AI services that run them only to perform the task you requested; it is processed and returned, not stored by hoot. See the Privacy Policy for details.

Voice cloning

When you use Voice Clone, you confirm you have the right to use the voice in the sample you provide. You agree not to create audio that impersonates a real person without their consent, or that is unlawful, deceptive, or harmful. You are solely responsible for the audio you generate.

Limitations

hoot is a browser-based tool with inherent limitations. Very large files may not process reliably due to browser memory constraints. We recommend keeping source files under 150 MB for the best experience.

Liability

hoot is provided without warranty. We are not responsible for any data loss, audio corruption, or other issues that may occur during use. Always keep backups of your original files.

Changes

We may update these terms from time to time. Continued use of hoot constitutes acceptance of any changes.

Contact

Questions about these terms: hey@hoot.audio.

Your plan

Account

How pricing works

Feedback

Rate your experience with hoot

Your notes go straight to the team.

Contact

We'd love to hear from you. Email is the fastest way to reach us.

We usually reply within a day.
Batch Processing
No chain recorded yet — apply some processors first
Drop audio files here, or click Add Files
Each file will get the chain you've applied in the editor.

Owliver

BETA
How can I help with your audio?
Workflows
?
Shortcuts
Playback
Play / PauseSpace
Go to start↵
Editing
Undo⌘Z
Redo⌘⇧Z
Copy audio⌘C
Paste audio⌘V
Mute audio⌥M
Trim audio
⌥T⌘⌫
Crop to selection⌥C
Delete selection⌫
Fades, Save & Export
Save project⌘S
Fade in 50ms⌥I
Fade out 50ms⌥O
Export audio⌘E
Navigation & Waveform
Zoom to selection⌘↑
Zoom to fit⌘↓
Zoom in / out
+–
Vertical zoom in / out
⇧+⇧–
Move playhead
←→
Nudge selection end
⇧←⇧→
Add markerM
Previous / next marker
,.
Toggle loopL
Fullscreen waveform
FEsc
Draw selectionClick & drag
Adjust selectionDrag edges
Adjust gainDrag orange line
Adjust fadesDrag handles
Pan viewSpace + drag
Scrub audioDrag playhead
Toggle crosshairC
Toggle snap to zero-crossS
How to use hoot