Why Text to Speech Voices Differ by Device (and How to Add More)

The short answer

The Text to Speech tool does not bring its own voices. It uses your browser's built-in speech feature (the Web Speech API), which lists the voices installed on your device and, in some browsers, voices the browser itself provides. That is why the same page shows different voices on a Windows laptop, a Mac, an Android phone and an iPhone. To get more voices, add them in your operating system's speech or accessibility settings, then reload the page.

Where the voices come from

When the page loads, the tool asks the browser for its list of voices and shows them in the voice menu as "name (language code)", for example en-GB or hi-IN. It selects the first English voice it finds, or the first voice in the list if there is no English one.

The list depends on three things:

• Your operating system. Windows, macOS, iOS and Android each ship different voices and let you install extra ones.
• Your browser. Some browsers add their own voices to the list. For example, desktop Chrome often lists voices with "Google" in the name alongside the system voices. Other browsers show only the system voices.
• What you have installed. Language packs and downloaded voices appear only after they are installed.

Because speech is produced by your device, the text you type is not sent to Lazyblink. Some browser-provided voices may use an online service from the browser maker; system voices that are installed on your device work offline.

How to choose a voice in the tool

Open the Text to Speech tool and type or paste your text.

Open the voice menu and pick a voice. The code in brackets tells you the language and region: en-US is American English, en-IN is Indian English, hi-IN is Hindi.

Set Speed (0.5x to 2x) and Pitch (0.5 to 2).

Press Speak. Use Pause, Resume and Stop while it plays.

The bottom line shows the word count and a rough duration, estimated at 150 words per minute adjusted for your speed setting.

Match the voice to the text's language

A voice reads every word using its own language's pronunciation rules. An English voice reading Hindi text in Devanagari may skip words or spell them out, and a Hindi voice reading English sounds unnatural. Pick a voice whose language code matches the text.

Adding voices on Windows 11

Open Settings, then Time & language, then Speech.

Under Manage voices, select Add voices.

Choose the voices or languages you want and install them.

Close and reopen your browser, then reload the Text to Speech page.

If you do not see Add voices, install Windows updates first. Another route is Time & language, then Language & region: add a language and make sure the text-to-speech option is selected. You may need an administrator account to install voices.

Adding voices on iPhone and iPad

To choose or download system voices:

Open Settings, then Accessibility, then Spoken Content.

Tap Voices and choose a language. Voices that are not yet downloaded show a download option.

Some voices come in an enhanced quality version. Apple notes these can be 100 MB or larger, so download them over Wi-Fi.

Then close and reopen Safari (or your browser) and reload the page. Browsers on iPhone and iPad use the system's speech engine, so the voices available to the tool follow what is installed on the device.

On a Mac or Android phone

• Mac: speech voices are managed in System Settings under Accessibility, in the spoken content section. Downloaded voices become available to Safari and other browsers that use system voices.
• Android: voices come from the device's text-to-speech engine, usually found in Settings under Accessibility or under language and input, depending on the phone maker. From there you can install voice data for more languages.

Menu names vary by version and manufacturer, so if a path above does not match your device, search your settings for "text-to-speech" or "spoken content".

Troubleshooting

The voice list is empty or shows only one voice

Some browsers load voices a moment after the page opens. The tool listens for the browser's "voices changed" signal and refreshes the list, so wait a second and open the menu again. If it is still empty, reload the page. On a freshly installed system, there may simply be only one voice until you add more.

"Web Speech API not available"

This message means your browser does not support speech synthesis. Try a current version of Chrome, Edge or Safari. Some privacy-focused browsers and in-app browsers (the ones that open inside social media apps) turn the feature off.

A new voice does not appear

Voices are read when the browser starts. Close every window of the browser, not just the tab, then open it again.

Speech stops partway through a long text

Some voices, especially online ones, can stop on very long passages. Split the text into sections of a few paragraphs and play them one at a time, or switch to a voice installed on your device.

Speed or pitch changes do nothing

A few voices ignore pitch, and some limit the speed range. Try another voice. Also note that the new setting applies the next time you press Speak, not to speech that is already playing.

The text is read with the wrong accent

Check the language code of the selected voice. An en-US voice will say "schedule" and "tomato" the American way. Choose en-GB, en-IN or another region if you prefer.

Can I save the speech as an MP3?

Not with this tool. The browser's speech feature plays audio through your speakers but does not give web pages the audio as a file, so there is no download button. If you need an audio file, use a dedicated recording or text-to-speech program on your computer.

Tips for natural-sounding speech

• Use punctuation. Commas and full stops create pauses. A long sentence with no punctuation is read in one breath.
• Write out symbols. "₹500" or "50%" may be read differently by different voices. "500 rupees" and "50 percent" are predictable.
• Expand abbreviations the voice might spell out, such as "approx." or "govt."
• Keep speed between 0.8x and 1.2x for listening practice or proofreading. Higher speeds are useful for skimming familiar text.

Related tools and guides

For using text to speech for study and proofreading, see How to Use Text to Speech Online. To go the other way and type by voice, use Speech to Text, and if your microphone is not picked up, read Speech to Text Not Working in Your Browser?. To check how long your script is before recording, paste it into the Word Counter.