Skip to content
WevoLabs

Multi-speaker · 100% free · Unlimited · No login

Free Zulu multi-speaker text to speech with multiple voices.

Two speakers or a whole cast: write the script as a conversation — Narrator:, Sara: — pick a voice for each name and get one finished MP3. Unlimited, no login, no hidden charges. Up to 8 speakers from 322 voices.

0 / 5,000 —

Lines

Rendered in order, each with its speaker’s voice.

    Write a script above — each speaker turn shows up here.

    Cast

      Lines you don’t change are reused, so edits re-render only what changed.

      Script format

      Names, colons, done.

      The studio reads the script top to bottom. A line that starts with a name and a colon begins that speaker’s turn; anything after it belongs to the same speaker until the next name.

      • Names can be anything — “Narrator”, “Dr. Rao”, “Customer”.
      • Add pauses inside a line with [[p700]].
      • Mix languages: each speaker can use a voice in a different language.
      • Edit one line and only that line is rendered again.
      Examplepodcast-cold-open.txt
      Narrator: It was 4 a.m. when the call came in. [[p700]]
Maya: Hello? Who is this?
Caller: You don't know me. But I know about the flight.
Maya: What flight?
Narrator: And that is how the strangest week of her life began.

      Three speakers, five lines, one MP3 — with a 0.7-second pause after the first sentence.

      Good for

      Scripts with more than one voice in them.

      Podcast segments

      Scripted intros, cold opens and two-host explainers.

      Stories & audio drama

      A narrator plus characters, each with a distinct voice.

      Language lessons

      Native-accent dialogues for listening practice.

      Training role-plays

      Customer and agent conversations for onboarding.

      Writing your first dialogue? Read the practical guide to multi-speaker AI dialogue. Narrating slides with one voice? See text to speech for presentations.

      Two voices or eight

      Text to speech with multiple speakers.

      Most single-voice tools read everything in one voice, so a dialogue sounds like one person talking to themselves. Here every speaker keeps their own voice from the first line to the last, and the lines are joined into one audio file with a short, natural gap.

      In other words, it is a free multi-speaker AI voice generator: you write the dialogue, and it generates the audio for every speaker — a two-person conversation, a story with several characters or a whole podcast segment.

      This is the studio that used to be called the Multi Narrator Studio — same address, new name.

      Two speakers

      A conversation between two people: an interview, a two-host podcast intro, a customer and an agent, or a language-lesson dialogue.

      Multiple character voices

      A narrator plus characters for stories, audio drama, games and audiobooks — each character easy to tell apart.

      Podcasts with multiple voices

      Turn a written episode into audio with a host, a guest and a narrator, then add music under the whole thing.

      Voices in different languages

      An English host and a Spanish guest, or a teacher and a student in two accents — each speaker can use any of the voices.

      Step by step

      Make a conversation in four steps.

      1. 01

        Write or paste the script

        Start every turn with a name and a colon. Lines without a name continue the previous speaker. Press “Example” to see a finished story.

      2. 02

        Check the cast

        Every name appears in the cast. Pick a voice for each one and, if you like, its own speed, pitch and volume.

      3. 03

        Generate

        Lines are rendered one by one with their speaker’s voice and joined with the gap you chose. Edit a line and only that line is made again.

      4. 04

        Listen and download

        Play the conversation, add background music if it needs it, then download MP3 or WAV.

      Better dialogue

      Make voices sound like people talking to each other.

      Contrast the voices

      Pair voices that are easy to tell apart — a different gender, accent or pitch — so listeners always know who is speaking.

      Keep turns short

      Real conversations move in short turns. Break long speeches up with a reaction from the other speaker.

      Let a narrator carry the scene

      For stories, a narrator line between dialogue sets the place and the mood, as in the example story.

      FAQ

      Multi-speaker questions.

      Can I make text to speech with two voices?
      Yes. Write the script as a conversation between two people — Sara: … and Adam: … — and give each name a voice. Two speakers is the most common setup; you can add up to 8.
      Is this the WevoLabs Multi Narrator Studio?
      Yes. The Multi Narrator Studio (Multi TTS) is now called the multi-speaker studio. It is at the same address, it is still free, and your earlier takes stay in your browser.
      How do I assign voices to speakers?
      Start each turn with a name and a colon — Host: Welcome back. Every name you use appears in the cast panel, where you choose its voice and fine-tune speed, pitch and volume. Lines without a name continue the previous speaker.
      My script has no names. Can I still use several voices?
      Yes. Switch “Split by” to paragraphs or sentences and the cast takes turns in order — useful for alternating narrators or for language-learning drills.
      How many speakers and how much text?
      Up to 8 speakers and 5,000 characters per take. For a longer episode, render it in parts; each take is saved in your browser and can be downloaded together as a ZIP.
      Why are some lines quicker the second time?
      Every rendered line is kept for the rest of your visit. When you edit one line or change one speaker’s voice, only the lines that changed are rendered again.
      Can voices speak different languages in one script?
      Yes. Give a speaker a voice in the language of their lines — for example an English host and a Spanish guest. Multilingual voices can also read text in several languages.
      Can I make a podcast or audiobook with multiple voices?
      Yes. A two-host podcast intro, an interview or a story with a narrator and characters all work the same way: one name per voice. Download the result as one MP3, or as WAV for editing.
      Can I add pauses or background music?
      Add [[p700]] inside any line for a pause, and set the gap between lines in the cast panel. After rendering, “Add music” mixes a track under the whole conversation in your browser.

      Need just one narrator?

      The single-voice studio is faster for voiceovers, lessons and articles.

      Single speaker Multi-speaker