About FreeTextoSpeech

Our Mission

FreeTextoSpeech started as a reaction to text to speech sites that ask for an email before you hear a single word. Here you paste text, pick a voice and download the audio. No account is needed for that, and every tool on the site works without one.

The people who use it most are creators making voiceovers for YouTube, TikTok and podcasts, students who take in reading better by ear, teachers preparing read-aloud material, and people with dyslexia or low vision who want a page read to them. The tools around the converter follow from those jobs: document import, dialogue and podcast studios, caption tools and a pronunciation fixer.

How It Works

The voices come from Kokoro, an open-weight speech model with 54 voices in 9 languages: US English (20 voices), UK English (8), Mandarin Chinese (8), Japanese (5), Hindi (4), Spanish (3), Brazilian Portuguese (3), Italian (2) and French (1). Cloud requests run on a GPU server we operate, and go to DeepInfra, which hosts the same model, when that server is busy. The English voices can also run entirely in your browser after a one-time model download, so the text never leaves your device.

Text is processed in memory and the audio is sent straight to your browser; we do not keep your text or the generated audio. Downloads are MP3 or 24 kHz WAV. We keep usage records (tool, character count, time) to apply limits, and remove the IP address from them after 90 days. The privacy policy lists every service that touches your data.

Why Free?

The free tier is meant to be enough for real work, not a trial. Without an account you get 25,000 cloud characters every 30 days, up to 5,000 per Generate. A free account raises that to 500,000 characters and 10,000 per Generate. The in-browser engine has no usage cap at all.

Running GPUs costs money, so free pages show ads. Optional paid plans (Ad-Free from $4 a month, Pro and Studio) remove the ads and add 1 to 10 million cloud characters a month. They do not unlock voices: all 54 voices, speed control and both download formats are on the free tier.

Commercial Use

You can use audio generated here commercially, on the free tier too: YouTube videos, podcasts, ads, courses, apps and games, with no attribution required. Kokoro's weights are released under the Apache 2.0 licence, which allows this. The full wording is in our Terms of Service .

Built By

FreeTextoSpeech is built and maintained by Bipul Kumar, an independent developer who writes the code and runs the servers. Some guides are drafted with the help of AI writing tools; facts such as limits, prices and voice counts are checked against the live tool and its code before publishing. How pages are written and corrected is described on the Team & Methodology page under About.