Skip to content
WKO AI

Kokoro · Speech generated on your device

Free text to speech in your browser

Hear a draft, practice a script, or make a short voiceover. Choose a voice and playback speed. Your text is processed locally, with no account or per-clip charge.

Turn text into speech

Create a speech clip in three steps

  1. Step 1

    Enter your text

    Open Speak and type or paste the words you want to hear.

  2. Step 2

    Choose a voice

    Select a voice and speed, then approve the first Kokoro model download.

  3. Step 3

    Listen and download

    Play the generated clip, regenerate it, or save the audio to your device.

Hear a Kokoro sample

“Welcome to the studio. Bring a notebook and a new idea.”

Generated with Kokoro, Heart voice, normal speed. Press Play to listen.

What to expect

Input is typed or pasted text; output is a playable, downloadable audio clip. The studio offers voice and speed controls, a local clip history, and regeneration. It does not clone voices or require you to upload a recording.

Kokoro’s speech model is approximately 86 MB, plus voice and runtime assets. The first use requires an internet connection and your download approval. Kokoro speech runs on the CPU through WebAssembly in your browser.

Pronunciation, names, and expression can vary. Start with a short paragraph and review the result. The current composer accepts up to 600 characters per clip. Longer passages take longer, especially on phones. Cached assets may support later offline use; test before disconnecting.

Need the reverse? Convert audio or video to text. You can also extract text from an image or read the ElevenLabs comparison.