Text to Speech

Turn any script into natural-sounding speech. Paste your script, pick a voice or clone your own from 60 seconds of audio, and get audio in seconds.

+ Your voiceExplore the voice library

100 / 1,000 · Free

  • 181,031,686 Videos generated
  • 156,809,347 Avatars generated
  • 26,685,426 Videos translated
Trusted by millions worldwide to bring their stories to life.
Features

Text-to-speech features powered by HeyGen Voice

Realistic AI voices from the #1 voice model

Type your text and HeyGen Voice reads it back with human pacing, emphasis, and inflection, a speech quality older engines never reached. Across 42,875 blind ratings in five languages, listeners chose it over every model tested, leading on clarity and native accent.

Natural AI voices for text to speech

Multilingual speech in 175+ languages, in your voice

Write once and generate speech in 175+ languages and dialects without paying voice actor fees in every market. HeyGen Voice keeps your voice across languages with a native accent, and listeners preferred it as narrator in 77.4% of decided votes in cross-language blind tests.

Text to speech in 175+ languages and dialects

Voice cloning from a 60-second recording

Record 60 seconds on any phone and AI voice cloning with HeyGen Voice learns your timbre, accent, and phrasing. For finer nuance, a Professional Voice Clone trains on 20+ minutes of your recordings to capture how your accent moves across words.

Voice cloning from a short sample

Control tone, speed, and delivery with text

Direct the read without touching a waveform. Voice Director sets one of 8 tones or follows your written direction. Set speed and volume per scene and add natural emphasis with Expressiveness Boost, so a launch sounds upbeat and conversational.

Text-based control over speech delivery

Speech that sounds recorded in the scene

Most AI speech sounds like a booth in every scene. HeyGen Voice reads the scene image and matches the room, mic distance, and delivery, from a quiet studio to a windy beach. The same script then flows into text to video with lip sync.

Text to speech presented on video by an AI avatar
Use cases

AI voice use cases for creators and teams

AI voiceovers for ads and marketing videos

Studio sessions cost thousands and take weeks to book. Generate an AI voiceover in your own cloned voice instead, and refresh the ad whenever the offer changes.

E-learning and course narration

Re-recording a course every time content changes stalls training teams. Edit the script, regenerate the lesson in the instructor's voice, and republish the same day with professional narration.

Audiobooks and podcasts

Mic setup and retakes leave many audiobooks and podcasts unfinished. Train a pro clone on your recordings, turn long-form written content into audio, and publish video versions with the audio to video converter.

Social media content for YouTube and TikTok

Content creators film clips in cars, parks, and on beaches, where studio-clean audio sounds wrong. HeyGen Voice matches each scene, so one voice fits every clip and content creation stays fast.

Accessible audio for readers with low vision

Text-only pages exclude people with visual impairments or dyslexia. Have articles and guides read aloud with clear AI narration, and add captions from the subtitle generator so every audience can follow along.

Dubbing videos for global markets

Traditional dubbing takes months of studio coordination per market. Translate a finished video's speech instead, keep the speaker's voice with a native accent, and match mouth movement so localized versions look native.

How it works

How to convert text to voice in four steps

Go from a written script to natural speech in your voice in four steps, inside HeyGen's browser-based video editor.

step icon

Add your script

Paste text from a doc or text file into AI Studio, or draft one with the script writer.

step icon

Choose or clone a voice

Filter voice options by language, accent, and style, or record 60 seconds to clone your own.

step icon

Fine-tune the delivery

Pick a tone, adjust speed, and preview lines for free until every one sounds right.

step icon

Generate and download

Click Generate, download your audio and SRT captions on paid plans, or export video up to 4K.

What is TTS?

TTS stands for text to speech, also called speech synthesis: software that converts written text into spoken audio. Modern TTS models are neural networks trained on human speech, so they predict pacing, intonation, and emphasis. On HeyGen's video and speech platform, that model is HeyGen Voice, which also reads the video scene so the audio fits the shot.

Do AI voices still sound robotic in longer content?

Not with current models, which produce realistic voices. In blind tests across 245 full-lesson scripts in five languages, listeners chose HeyGen Voice as narrator over every model it faced, winning 61.5% to 77.9% of decided votes. Voice quality holds from the first line to the last, and raising Expressiveness Boost fixes a line that reads flat.

How do I convert text to speech and download the audio?

Paste or type your script in AI Studio, choose a voice, and select Generate. Previews are free on every plan. On paid plans you can save the audio on its own with SRT captions, or keep it in the video with synced lip movement on an on-screen presenter in the AI video generator. Developers get a WAV file back from each request.

What is the best AI text to speech tool?

It depends on what the audio is for. A simple text reader is enough for reading documents to yourself. When other people will hear it in videos, ads, and courses, you need natural voices, so compare text to speech tools on blind listening tests. HeyGen Voice ranks #1 on Artificial Analysis, and the method is in the HeyGen Voice research report.

Can TTS replace recorded voiceovers for online courses?

Yes, creators are running full course businesses on generated audio. Educator Anton Voroniuk reports saving 15.5 hours per week and cutting production costs 40x after switching to HeyGen, while reaching more than 1 million students with AI-narrated lessons instead of studio recordings. The full numbers are in the Anton Voroniuk customer story.

Is there a free text-to-speech plan, and what does it cost?

Yes. The free plan needs no credit card and includes 3 videos a month up to 1 minute each, 1 voice clone, and free previews, so free users can test voices first. Paid plans start at $24 per month, the pro clone add-on costs $29 to $99 per month, and developers pay $30 per 1 million characters.

How many languages and accents does HeyGen support?

HeyGen generates speech in 175+ languages and dialects. The voices available in the AI voice generator library filter by language, accent, age, and gender, so a UK-accented explainer and a US-accented ad can share one script. HeyGen Voice keeps your cloned voice across supported languages with a native accent.

Can I clone my own voice for TTS?

Yes. Record or upload a clip from any device: 60 seconds works, and 1 to 3 minutes is ideal. HeyGen Voice learns your timbre, accent, and phrasing, usually within seconds. Voice Clone adapts delivery to each scene, while Full Clone preserves your recorded session. Free plans include 1 clone, Creator 40, and Pro 70.

Can I use the generated speech commercially?

Yes. HeyGen allows commercial use of what you create, as long as your inputs follow its usage guidelines, so the audio can run in ads, client campaigns, paid courses, and product launches. Clone only voices you own or have permission to use.

Can I control tone, pauses, and pronunciation?

Yes. Voice Director offers 8 tones, including Calm, Excited, and Serious, plus free-text direction. Voice Mirror lets a library voice perform a line with your pacing and emotion, and you can compare different voices on the same line. For brand names, spell them phonetically in the script to steer pronunciation.

Is there a character limit for long scripts?

The editor splits long scripts into segments of up to 1,000 characters, so you build long narration scene by scene instead of hitting one cap on the whole script. On the free tier, videos cap at 1 minute. Developer requests take up to 5,000 characters each.

Should I download voiceovers as MP3 or WAV?

WAV is uncompressed, so it suits editing, mixing, and archiving. MP3 is compressed, so the audio file is smaller and plays anywhere, which suits sharing. HeyGen Voice returns WAV to developers, and any audio editor can convert it to MP3 when file size matters.

Is there a TTS API for developers?

Yes. The HeyGen Voice API takes a voice ID, text, and language in one REST request and returns a WAV file URL. You can also stream audio over SSE with word timestamps for captions. Pricing is $30 per 1 million characters, with no packages or minimums.

Is my text and voice data private?

HeyGen excludes enterprise customer data from model training by default, and other customers can opt out. Details on how data is stored and protected are on the HeyGen security page.

Start creating with HeyGen Voice

Turn any script into human-like speech in your own voice, matched to every scene.