Text to Speech Preview
Listen to any text read aloud in the browser — free, instant, no signup.
Updated
What is the Text to Speech Preview?
Text to Speech Preview turns any written text into spoken audio — instantly, in your browser, with no account or installation required. Powered by the Web Speech API, it uses the same high-quality voices already built into your operating system: the same ones Siri, Narrator, and VoiceOver rely on.
Paste a paragraph, an email draft, a blog post, or a foreign-language phrase, pick a voice, and hit Play. Every word lights up as it is spoken so you can follow along and catch the exact moment something sounds off.
Whether you are proofreading copy, checking pronunciation, estimating read time, or simply want to rest your eyes and listen, this tool handles it with no file uploads and no data ever leaving your device.
How it works
Voice loading
On first load the tool asks the browser for its list of available SpeechSynthesisVoice entries. Chrome and Edge load voices asynchronously and fire an onvoiceschanged event when they are ready; Firefox and Safari return them synchronously. The tool handles both cases and populates the Language and Voice dropdowns automatically once voices are available.
Voices marked with ☁ are streamed from the operating system's cloud service and require an internet connection. Voices without that marker are stored locally and work fully offline. The system default voice is marked with ★.
Tokenisation and word highlighting
When you press Play the tool splits your text into a flat list of word tokens and gap tokens (spaces, punctuation, newlines), each carrying its start offset in the original string. The Web Speech API fires a boundary event for every word it speaks, reporting a charIndex — the position of that word in the utterance. The tool maps that offset back to the correct token and highlights the matching <span> in the Now Reading panel, scrolling it into view automatically.
Sentence navigation
The tool pre-scans the text for sentence boundaries (., !, ? followed by whitespace) and stores each sentence's start offset. The ⏮ skip-back button jumps to the beginning of the current sentence (or the previous one if you are within the first eight characters), and ⏭ skip-forward jumps to the next. Both work by cancelling the current utterance and starting a new one from the target offset, keeping the highlight panel in sync.
Chrome long-text workaround
Chromium-based browsers silently pause speechSynthesis after roughly 15 seconds when the tab is in the background. The tool runs a background interval that calls speechSynthesis.resume() whenever it detects a paused state, keeping long texts playing uninterrupted.
Privacy
All processing happens entirely in your browser. Your text is never sent to any server — not even Anthropic's. The only network activity that can occur is the OS streaming a ☁ cloud voice directly from your operating system vendor (Apple, Google, or Microsoft) if you have selected one.
Controls at a glance
| Control | What it does |
|---|---|
| Language | Filters the voice list to a single language or accent group |
| Voice | Selects the specific synthetic voice to use |
| Speed | Multiplies the natural speaking rate (0.25× to 4×) |
| Pitch | Adjusts the fundamental frequency of the voice (0 to 2) |
| Volume | Sets output loudness (0 % to 100 %) |
| ⏮ / ⏭ | Jump to the previous or next sentence during playback |
| Stop | Cancels the current utterance and resets the highlight |
Examples
Proofreading an email
Paste a draft email and play it back to catch awkward phrasing your eyes skip over.
Previewing a blog post introduction
Test how the opening paragraph of an article sounds before publishing.
Checking a foreign-language phrase
Select a Spanish or French voice to hear how a translated sentence is pronounced.
Estimating podcast script length
Paste a script and read the estimated duration stat to check it fits your target slot.
Frequently asked questions
Does this tool send my text to any server?
Does this tool send my text to any server?
No. Everything runs entirely inside your browser using the built-in Web Speech API. Your text is never transmitted to Toolection's servers or to Anthropic. The only exception is if you select a ☁ cloud voice — in that case your operating system streams the audio directly from the voice provider (Apple, Google, or Microsoft), exactly as it would if you used your device's built-in screen reader.
Why is the voice list empty or not loading?
Why is the voice list empty or not loading?
Chrome and Edge load voices asynchronously. If you open the tool and the dropdowns appear empty, wait one or two seconds for the browser to fire its voiceschanged event and the list will populate automatically. If voices never appear, check that your operating system has at least one text-to-speech voice installed — on Windows this is under Settings → Time & Language → Speech; on macOS under System Settings → Accessibility → Spoken Content.
What is the difference between a local voice and a cloud voice (☁)?
What is the difference between a local voice and a cloud voice (☁)?
Local voices are installed on your device and work with no internet connection. Cloud voices are streamed in real time from your OS vendor's servers (Apple, Google, or Microsoft) and produce higher-quality, more natural-sounding audio. Cloud voices are marked with ☁ in the voice picker. If you are working with sensitive text, choose a local voice to keep everything fully offline.
Can I use this tool to preview text in a language other than English?
Can I use this tool to preview text in a language other than English?
Yes. Use the Language dropdown to filter voices by language or accent. If your operating system has voices installed for that language — or if cloud voices are available — they will appear in the Voice dropdown. The number and quality of available languages depends entirely on what your OS and browser support.
Why does playback stop on its own after a few seconds in Chrome?
Why does playback stop on its own after a few seconds in Chrome?
This is a known bug in Chromium: when the browser tab is in the background, speechSynthesis silently pauses after roughly 15 seconds. This tool works around it automatically by checking for a paused state every 10 seconds and calling resume() — so long texts should play through without interruption. If it still stops, try keeping the tab in the foreground.
Do speed, pitch, and volume changes apply immediately while something is playing?
Do speed, pitch, and volume changes apply immediately while something is playing?
No. Changes to Speed, Pitch, Volume, and Voice take effect the next time you press Play or Resume. A note is shown below the controls while audio is active as a reminder. This is a limitation of the Web Speech API — the properties of an utterance cannot be changed once it has started.
How is the estimated duration calculated?
How is the estimated duration calculated?
The tool counts the number of words in your text and divides by a baseline rate of 150 words per minute — a common average for English text-to-speech at Speed 1×. It then divides again by your current Speed setting. The figure is a rough guide; actual duration varies by voice, language, and how much punctuation-driven pausing the voice applies.
Why does the tool warn me about long texts?
Why does the tool warn me about long texts?
Chromium-based browsers impose an internal character limit on a single SpeechSynthesisUtterance (roughly 32 000 characters, but reliable behaviour drops off well before that). For texts over 5 000 characters the tool shows a warning. If your text is long, consider splitting it into sections and playing each one separately.
Can I download the audio as an MP3 or WAV file?
Can I download the audio as an MP3 or WAV file?
Not currently. The Web Speech API does not expose a way to capture the audio output as a file — the audio is routed directly to the device's speakers. For downloadable audio you would need a dedicated TTS service such as Google Cloud Text-to-Speech, Amazon Polly, or ElevenLabs, which generate audio files server-side.
What browsers are supported?
What browsers are supported?
The tool works in Chrome, Edge, and Safari, which all implement the Web Speech API with a solid selection of voices. Firefox has partial support but may offer fewer voices depending on the operating system. The tool shows an error message if your browser does not support speech synthesis at all.
Related tools
More utilities you might find handy.