--- title: "Multi-voice text to speech: dialogues in one MP3 | Speecho" description: "Paste a script with names and every character gets its own AI voice. Up to 8 voices in one MP3, priced per character like any other text. No subscription." canonical: "https://speecho.app/multi-voice-text-to-speech/" language: "en" publisher: "Speecho" --- Text to speech · Dialogue # Multi-voice text to speech: a voice for every character. Paste a script where each line starts with a name, like "Anna: Did you book the table?", and every name becomes a character with its own voice. Speecho reads the lines in order, with the pause you choose between them, and hands back one MP3. Up to 8 characters and 100 lines, priced per character like any other text. [Voice your first dialogue →](https://my.speecho.app/voice?mode=dialogue) Same price as the same text in one voice · credits never expire · no subscription. Every voice has a free preview. [Watch the full tutorial on YouTube](https://www.youtube.com/watch?v=SMr6w-GqSVM) ## Hear a dialogue first Voiced in the app from the scripts below, on Premium voices, with a normal pause between lines. ### Two friends and a forgotten booking Voices: **Anna** (Jessica) , **Tom** (Mark) 0:21 1. **Anna:** Did you book the table for Friday? 2. **Tom:** \[sighs] I forgot. Again. 3. **Anna:** Tom! It's my sister's birthday. 4. **Tom:** I know, I know. I'll call them right now. 5. **Anna:** \[laughing] Too late. I booked it this morning. 6. **Tom:** Then why did you ask? 7. **Anna:** I wanted to see your face. ### A Spanish lesson: ordering a coffee Voices: **Teacher** (Hope) , **Barista** (Carolina) , **Customer** (David) 0:22 1. **Teacher:** Listen to how the barista greets you. 2. **Barista:** ¡Buenos días! ¿Qué le pongo? 3. **Teacher:** She is asking what you would like. Answer like this. 4. **Customer:** Un café con leche, por favor. 5. **Barista:** ¿Para tomar aquí o para llevar? 6. **Teacher:** For here, or to take away? Say: para llevar. 7. **Customer:** Para llevar, gracias. ### Two colleagues, in Japanese Voices: **さくら** (Morioki) , **けんた** (Hinata) 0:20 1. **さくら:** 明日の会議、何時からだっけ？ 2. **けんた:** 十時からだよ。資料はもう送った？ 3. **さくら:** \[sighs] まだ。今夜やるね。 4. **けんた:** 手伝おうか？ グラフなら得意だよ。 5. **さくら:** 本当？ じゃあお願い。お礼にコーヒーおごるね。 6. **けんた:** \[laughs] それなら二杯で。 ## From a script to one MP3 No recording, no audio editor, no stitching clips together by hand. friday-dinner.mp3 2 characters · 7 lines · Premium · normal pause Ready 1 **Paste the script.** One turn per line, with a name and a colon in front. Paste it into Text to speech and the app offers to switch to Dialogue, or build it line by line in the Dialogue view. 2 **Check the cast.** Every name gets a voice that speaks the language of its lines. Swap any of them, hear every voice free, and pick the pause between lines. 3 **Voice it and download.** The lines are voiced in order and joined into one MP3. Change a line later and only that line is voiced again. ## What people make with it Anything written for more than one voice. ### Language lessons A listening exercise needs two speakers, and the textbook audio never matches your lesson. **Now:** two native voices for your own dialogue, or a teacher in English and a waiter in Spanish in one file. ### Podcast intros and ads A two-host intro or an ad read means booking two people for thirty seconds of audio. **Now:** two voices read the script as written, ready to drop into your edit. ### Explainer videos One narrator doing every role flattens a story with two characters. **Now:** a voice per character, with directions like \[whispering] or \[laughs] typed into the line. ### Role-play training Sales and support role-plays are read off a page, so trainees never hear the tone. **Now:** the customer and the agent voiced like a real call, line by line. ### Game and app prototypes Placeholder dialogue waits weeks for voice actors. **Now:** every line voiced in minutes, and redone line by line when the script changes. ### Audio stories You read every character yourself, or the story stays on the page. **Now:** up to eight characters, each with a voice of its own, in one file. ## A podcast from any text Paste an article, a chapter or your notes, or open a PDF, Word or EPUB file, and Make a podcast from text turns it into a conversation between two hosts of about 5, 10 or 20 minutes. The hosts stick to your text. The script opens in the Dialogue editor, where you check the lines and swap the voices, then it is voiced into one MP3 like any dialogue. Writing the script costs 15 credits for up to 5 minutes, 25 for up to 10, 40 for up to 20, plus 1 credit per 3,000 characters of text. [Text to podcast →](https://speecho.app/text-to-podcast/) ## What a dialogue can hold - 8 characters, each with a voice of its own - 100 lines, up to 5,000 characters each - 4 pause lengths between lines: none, 0.3, 0.6 or 1 second - \[ ] voice directions in any line: \[whispering], \[laughs], \[pause] - 31 languages with native voices, and each character can speak a different one - 1 MP3 for the whole dialogue, and only changed lines are voiced again ## Priced like any other text A dialogue costs what the same text costs in one voice: per 1,000 characters, by tier. More voices never cost more. Standard Clear, everyday use 50 credits / 1,000 characters ≈ $0.10 / 1k Standard voices: clear and natural, for drafts, lessons and internal audio. Enhanced ElevenLabs 110 credits / 1,000 characters ≈ $0.23 / 1k ElevenLabs: more expressive, the sweet spot for published dialogue. Premium ElevenLabs 160 credits / 1,000 characters ≈ $0.33 / 1k ElevenLabs: the most natural, with native speakers at their best. ≈ 1,000 characters make about one minute of speech. One tier per dialogue; the exact price shows on the button before you voice it, and the smallest charge is 10 credits. ## Frequently asked questions **How do I make text to speech with several voices?** Open Text to speech in the app and switch from Text to Dialogue, or paste a script where each line starts with a name and a colon: the app offers to switch for you. Every name becomes a character with its own voice. Change any of them, then press Voice dialogue. You get one MP3 with all the lines in order. **How much does a dialogue cost?** Exactly what the same text costs in one voice: per 1,000 characters at the rate of the tier (50 credits on Standard, 110 on Enhanced, 160 on Premium), with a minimum of 10 credits. The number of speakers and voices does not change the price, and the total shows on the button before you voice anything. **How long can a dialogue be?** Up to 8 characters and 100 lines, with up to 5,000 characters in a line. A long dialogue is voiced in portions: the first lines, about 5,000 characters, right away, and the rest after you have heard them and decided to go on. **Can the characters speak different languages?** Yes. Each character has a voice of its own, and each voice reads the language of its lines: a teacher in English and a waiter in Spanish, or two native speakers of any of the 31 languages with native voices. All voices in one dialogue come from the same tier. **Can I change one line without redoing the whole dialogue?** Yes. Open the dialogue from your history, edit the lines and voice it again. Unchanged lines keep their audio, and only the changed ones are voiced and charged. Or press Voice this line again on one line: pick another voice, fix a word or add a direction like \[whispers], and only that line is voiced anew. The same text costs the line's price again; a changed text costs the price of the new text. **Do voice directions work in a dialogue?** Yes, line by line: type \[whispering], \[laughs], \[pause] or another short direction in English inside the line. Enhanced and Premium voices take any short direction. The universal Standard voices read plain text, so their tags are removed before voicing. **What do I get at the end?** One MP3 of the whole dialogue, with the pause you chose between lines. It stays in your history for 90 days, where you can play it, download it again or edit it. **Can I get subtitles for a dialogue?** Yes, with Enhanced and Premium voices. The arrow next to the MP3 gives you an SRT or VTT file timed word by word to the audio, and Subtitles with names (SRT) puts each character's name in front of their lines, which suits language lessons and translated interviews. Subtitles cost nothing extra; Standard voices do not have them. **Can I use the audio commercially?** Yes. Generation runs on the paid commercial API tiers of our voice providers, whose terms permit commercial use of the generated audio. You are responsible for the content of the script. ## Give every character a voice. Paste your script and hear every voice free. You pay per character of text, however many voices read it. [Voice a dialogue →](https://my.speecho.app/voice?mode=dialogue) Updated October 2026