About this text to speech
Text to speech turns written words into spoken audio using the speech synthesis features provided by your browser and operating system. It can help you listen to a draft, check the rhythm of a sentence, practice pronunciation, or follow a passage while reading the original text. Paste a short passage, choose an available voice, and press Speak to start listening.
The voice list includes only voices marked as local by the browser. This avoids selecting a remote speech service for your text. Voice names, languages and availability vary between computers, phones and browsers; an empty list means that no compatible local voice is currently available. Installing an operating system voice may help.
Speed changes the reading rate, while pitch changes the apparent tone. Pause and Resume control the current reading; Stop cancels it. This tool does not create an audio download. For lengthy documents, read one section at a time, because browser speech engines may stop or impose their own length limits.
Which controls should I check first?
The Text to Speech form labels its controls as Text to read aloud; Local voice; Speed (0.5–2); Pitch (0–2). Use those labels to keep the source and settings separate. A sample argument that selects a mode is not additional source text. Choose the setting matching your intended operation before copying a result, and change it explicitly when comparing another mode. The labels describe what this form accepts; the worked examples below demonstrate only the stated supplied values.
How should I interpret this specific operation?
A usable local voice must be available for the selected language. Changing pitch or rate alters delivery rather than the text's meaning. A settings example verifies the parsed controls but does not make a speaker audible. Test a short sample with the desired voice before a long passage, and retain the source text because this page does not generate a persistent audio-file download.
Worked helper values
These values come from the reproducible calculation examples supplied with this tool. Each row identifies the complete supplied arguments and the returned value. A visible ↵ denotes a line break and ⇥ denotes a tab in text examples. Settings appear alongside source data so that a different mode is not mistaken for an identical input. These are helper calculations, not observations of a live device, fetched target, image decoder, browser preview or downloaded file. Use the stated method above to relate the returned fields to the actual operation.
| Supplied values | Calculated return value |
|---|---|
| Hello; 1.5; 1 | {"text": "Hello", "rate": 1.5, "pitch": 1} |
| Hello; 1; 0 | {"text": "Hello", "rate": 1, "pitch": 0} |
What counts as one text element?
Text can be handled as visible graphemes, Unicode code points, bytes, words or complete lines. These units differ. A joined family emoji may look like one character while containing several code points; an accented character may use more than one encoded byte. Use the unit stated in this tool's method before comparing results with an editor or database. Line tools should receive one entry per line, while word tools use their documented whitespace boundaries. A transformation that preserves every visible word can still change separators, and a transformation that preserves every byte does not necessarily operate on complete visible characters.
How should I prepare a representative sample?
Start with a short excerpt containing the features that matter to your job: an accented letter, a number, punctuation, a repeated item, or a blank line. Choose the relevant options and inspect the result before processing the full text. Keep a separate source copy so that you can compare exact spaces and case. If duplicate labels refer to different records, attach identifiers before applying a list operation that could remove them. Text processing cannot infer whether two identical-looking names belong to the same record. A small sample makes delimiter mistakes easier to see than a long unstructured paste.
Which differences should I inspect after conversion?
Compare the first and last entries, any blank entries, and text around punctuation. Check whether the intended operation preserves line endings, joins tokens with spaces, or substitutes a custom delimiter. These are distinct behaviors: reversing words does not preserve the original layout, while reversing lines keeps each line's internal content together. A normalized comparison can intentionally ignore case or accents, whereas an exact comparison retains them. Read the stated method instead of assuming that all text tools share one normalization policy. If another application reports a different count, compare the boundary rules before treating either number as incorrect.
Why can copied text behave differently elsewhere?
A destination can apply its own normalization, line-ending rules, font substitution or length limits. Fancy Unicode letters are different characters from ordinary ASCII letters even when they look like a font change. Some destinations cannot display those glyphs, and searching plain text may not match them. Likewise a byte count changes with encoding, while a visual count may remain unchanged. Test a small converted sample in the receiving application before replacing a large document. Retain ordinary text when accessibility, searching or exact identifiers matter. A local text result is data rather than proof that the destination interprets it the same way.
Why is the voice list empty?
Your browser has not exposed a local voice. Wait briefly, install a system voice, or try another browser.
Can I download an MP3?
No. Web Speech plays audio through your device; this tool does not record or export it.
Does this use a remote speech provider?
Only voices marked localService by the browser are offered. Remote voices are excluded.