Text to speech API

A text-to-speech API with 57 voices in 14 languages, streaming audio and $3 per million characters. Your first 100,000 characters are free every month, with no card, and the fastest voices reach first audio in about 44 ms (our measurement). Send up to 500 characters, get back a link to the audio; lines already in the checked library come back instantly and any other text is generated on demand. Billing is per character with no subscription.

Text to speech API at a glance

Speakvora text to speech API
Price$3 per million characters for catalog voices, $3 for your own voice. Compared with Polly, Google, OpenAI and ElevenLabs
Free100,000 characters per account, no card. Free tiers compared
Voices57 live voices in 14 languages, including 156 English stock voices. Hear them
SpeedPhonon voices reached the first audible sound in 44 ms (median, 95th percentile 66 ms) in our November 21, 2026 test from AWS us-east-1, level with Amazon Polly Standard (43 ms) and about three times faster than Polly Neural (144 ms). Method and every number
Limits1 to 500 characters per request; 15 requests per second on the free plan, 100 paid. No SSML or viseme marks. Word and sentence speech marks are supported (marks: true), with word times measured from the audio.

First request

cURL

curl -X POST https://speakvora.com/api/v1/speak \
  -H "x-api-key: $SPEAKVORA_KEY" \
  -H "content-type: application/json" \
  -d '{"text":"Your order is ready for pickup.","voice":"af_heart","lang":"en"}'

# 200 {"url":"https://speakvora.com/voice/v2/af_heart/..../r1.m4a","lang":"en","voice":"af_heart","characters":32,"format":"m4a"}

JavaScript (server)

// Run this on your server (Node 18+). Never put the key in browser code.
const r = await fetch("https://speakvora.com/api/v1/speak", {
  method: "POST",
  headers: { "x-api-key": process.env.SPEAKVORA_KEY, "content-type": "application/json" },
  body: JSON.stringify({ text: "Your order is ready for pickup.", voice: "af_heart", lang: "en" }),
});
if (!r.ok) throw new Error((await r.json()).error);   // e.g. unauthorized, free_allowance_used, rate_limited
const { url } = await r.json();                        // link to the audio file

Python (requests)

import os, requests

r = requests.post(
    "https://speakvora.com/api/v1/speak",
    headers={"x-api-key": os.environ["SPEAKVORA_KEY"]},
    json={"text": "Your order is ready for pickup.", "voice": "af_heart", "lang": "en"},
    timeout=30,
)
r.raise_for_status()
print(r.json()["url"])   # link to the audio file

Responses are JSON with a url. Fetch that URL to download the audio. Retry 502, 503 and 504 once for brand-new sentences. New to this? Read what a text-to-speech API is.

Two ways to get speech

Library (catalog) speech

A checked sentence that already exists as audio. The API returns a link to the stored file, so there is no generation delay. Search the library with /v1/search.

Live generation

Any other text. It is synthesised on first request, then reused. Brand-new sentences took about 1 second in our latest test with a warm GPU worker, and several seconds without one (see the speed table). Plan for both.

Voices and languages today

Speakvora has 57 live voices in 14 languages, including 156 English stock voices you can hear at /voices. Phonon voices start speaking in about 44 ms (our own measurement; your network adds its own delay), and your first 100,000 characters are free with just a sign-up. The catalog voices below cover 14 languages (English, Arabic, Danish, German, Spanish, Finnish, French, Hindi, Italian, Japanese, Korean, Portuguese, Swedish, Chinese); industry agents exist in 13 languages. Live generation is offered for our fast catalog voices; the language you send must match the voice. Numbers: English accepts digits, other languages need numbers written as words. Non-English results carry a "native review pending" label until a speaker approves them.

VoiceIdTierDescriptionChecked clipsStatus
Easygoing Friendeasygoing-friendPremiumRelaxed, natural and friendly. One identity across many languages.en: 802, de: 34, es: 31, fr: 29, ja: 27, ko: 33, pt: 29, zh: 27library only
Arabic voicechatterbox-arChatterboxClear, natural Arabic voice.ar: 2,260library only
Danish voicechatterbox-daChatterboxClear, natural Danish voice.da: 395library only
German voicechatterbox-deChatterboxClear, natural German voice.de: 965library only
Finnish voicechatterbox-fiChatterboxClear, natural Finnish voice.fi: 2,080library only
Korean voicechatterbox-koChatterboxClear, natural Korean voice.ko: 827library only
Swedish voicechatterbox-svChatterboxClear, natural Swedish voice.sv: 83library only
Deliaph_deliaFastLow, grounded and sincere. Her apologies sound honest.live
Eastonph_eastonFastPunchy and close in, with crisp consonants that cut through a noisy line.live
Zoeyph_zoeyFastPlayful and upbeat, American.live
Juniperph_juniperFastCheerful with a warm, resonant core, American.live
Everlyph_everlyFastSoft and relaxed, with an easy, unhurried delivery. American.live
Harperph_harperFastModern, confident and friendly, American.live
Hadleyph_hadleyFastEven-toned and kind, with crisp consonants that carry over a phone line. American.live
Micahph_micahFastBright and energetic, with a playful lift. American.live
Russellph_russellFastHigh-energy and encouraging. American.live
Prestonph_prestonFastFocused and steady, with clean consonants. American.live
Milesph_milesFastResonant and decisive, with real dynamic range. American.live
Maeveph_maeveFastSparky and attentive, gets to the point with a smile. British.live
Pippaph_pippaFastConfident and upbeat, with infectious energy. British.live
Alfieph_alfieFastPunchy and forward, with crisp consonants. British.live
Edwinph_edwinFastEfficient and firm, with tidy articulation that holds up on a phone line. British.live
Tobyph_tobyFastSparky and friendly, quick to help. British.live
Teddyph_teddyFastQuick and close in, with firm consonants and a wide pitch range. British.live
Morwennaph_morwennaFastCharacter voice: low, husky and slow, close to the mic.live
Old Mortlakeph_mortlakeFastCharacter voice: an elderly gravedigger, low, raspy and half whispered.live
Fennickph_fennickFastCharacter voice: a curious, excited boy.live
The Thing Under the Stairsph_understairsFastCharacter voice: a gravelly, near-floor growl.live
Vyrmathraxph_vyrmathraxFastCharacter voice: an ancient dragon, very low, slow and dry.live
The Snow Queenph_snowqueenFastCharacter voice: cool, poised and precise.live
Sir Bramwellph_bramwellFastCharacter voice: a booming, pompous English knight.live
Fairy Godmotherph_fairygodmotherFastCharacter voice: a warm, velvety grandmotherly alto.live
Ivypk_ivyFastHigher voice, energetic delivery.live
Lenapk_lenaFastHigher voice, balanced delivery.live
Marisolpk_marisolFastHigher voice, balanced delivery. Spanish-accented English.live
Norapk_noraFastMid voice, energetic delivery.live
Piperpk_piperFastHigher voice, balanced delivery.live
Quinnpk_quinnFastHigher voice, energetic delivery.live
Sashapk_sashaFastHigher voice, energetic delivery.live
Antonpk_antonFastDeep voice, balanced delivery.live
Arunpk_arunFastMid-low voice, energetic delivery.live
Brampk_bramFastDeep voice, calm delivery.live
Colepk_coleFastDeep voice, energetic delivery.live
Dalepk_daleFastMid-low voice, balanced delivery.live
Devpk_devFastMid voice, energetic delivery.live
Edmundpk_edmundFastDeep voice, energetic delivery.live
Emilpk_emilFastDeep voice, energetic delivery.live
Enzopk_enzoFastDeep voice, balanced delivery. Spanish-accented English.live
Felixpk_felixFastMid-low voice, balanced delivery.live
Guspk_gusFastMid-low voice, energetic delivery.live
Hankpk_hankFastMid-low voice, balanced delivery.live
Hugopk_hugoFastDeep voice, calm delivery.live
Idrispk_idrisFastDeep voice, balanced delivery.live
Jasperpk_jasperFastMid-low voice, calm delivery.live
Kianpk_kianFastMid-low voice, calm delivery.live
Mateopk_mateoFastMid voice, calm delivery. Spanish-accented English.live
Milopk_miloFastMid-low voice, energetic delivery.live
Ottopk_ottoFastDeep voice, energetic delivery.live
Ramonpk_ramonFastMid-low voice, balanced delivery.live
Reedpk_reedFastDeep voice, energetic delivery.live
Rowanpk_rowanFastMid-low voice, energetic delivery.live
Silaspk_silasFastDeep voice, energetic delivery.live
Theopk_theoFastMid-low voice, energetic delivery.live
Waltpk_waltFastMid-low voice, balanced delivery.live
LanguageCodeLibrary clipsReview
Englishen7,730approved
Arabicar2,260approved
Danishda395approved
Germande965approved
Spanishes12,547approved
Finnishfi2,080approved
Frenchfr13,203approved
Hindihi10,666approved
Italianit10,442approved
Japaneseja11,198approved
Koreanko827approved
Portuguesept13,113approved
Swedishsv83approved
Chinesezh6,013approved

Price

WhatList priceUnit
Catalog voices$3per 1,000,000 characters
Your own voice$3per 1,000,000 characters
Free allowance100,000 charactersper account, no card; covers any catalog-voice text, new sentences included

Library lines are billed by character like any other text, and a repeat of the same sentence is billed again. List prices. See pricing and comparisons.

Speed

Phonon voices reach the first audible sound in about 44 ms and Pocket voices in about 90 ms (measured 21 November 2026 with the client in AWS us-east-1 over the public internet; your own network adds its own delay). Library and prepared phrases come back instantly. See the measured tables, the method and the caveats on the speed page and the streaming guide. These are our measurements, not guarantees.

Latest head-to-head, November 21, 2026, Coval's method, client in AWS us-east-1, every service over the public internet, median time to the first audible sound: Speakvora Phonon 44 ms, Pocket 90 ms, Amazon Polly Standard 43 ms, Polly Neural 144 ms, Polly Generative 186 ms. Full table and method.

Where the speed figure comes from: it is our own vendor-run test (November 21, 2026), not an independent benchmark. The rerunnable method and the raw data are in the public tts-latency-benchmark repository, and every price we compare is in the open TTS price index. Speed figures that other vendors publish, for example Deepgram or Cartesia, use different methods and are not directly comparable with ours.

Amazon Polly Standard and Google Cloud Standard at $4, next to Speakvora at $3 per million characters

Speakvora lists $3 per million characters for catalog voices, with 100,000 characters of speech free every month (renewing on the 1st, UTC) and no card to start. Amazon Polly Standard lists $4 per million characters, with its own free tier of 5 million Standard characters a month. Google Cloud Text-to-Speech Standard voices list $4 per million characters, with 4 million free characters a month. These are list prices only; the voices, models and languages differ, so this is a cost comparison, not a quality comparison.

Provider and tierPrice per 1M charactersFree tier, as that provider's page states itRead onSource
Speakvora catalog voices$3100,000 characters of speech and 60 minutes of speech to text every month, renewing on the 1st (UTC), no card22 November 2026this site
Amazon Polly Standard$45 million Standard characters a month, no time limit stated22 November 2026source
Amazon Polly Neural$161 million characters a month for the first 12 months22 November 2026source
Amazon Polly Generative$30100,000 speech requests a month for the first 12 months22 November 2026source
Amazon Polly Long-Form$100500,000 characters a month for the first 12 months22 November 2026source
Google Cloud Standard$44 million characters a month22 November 2026source
Google Cloud Neural2$161 million characters a month22 November 2026source
Google Cloud Chirp 3 HD$301 million characters a month22 November 2026source
Google Cloud Studio$1601 million characters a month22 November 2026source

What is different: Polly and Google are general cloud platforms with many voices and languages and a wider set of cloud services around them; Polly bills speech marks like speech. Speakvora has 177 live voices in 14 languages, returns speech marks when you ask for them (marks=true in the API), plays library lines from stored audio, and bills your own voice at the same $3. If you are inside Polly's 5 million free Standard characters a month, Polly costs less than us for that volume. Polly prices read 22 November 2026 on aws.amazon.com/polly/pricing; Google prices read 22 November 2026 on cloud.google.com/text-to-speech/pricing. Prices change; confirm at the source.

Cheaper list prices exist: DeepInfra lists $0.99 and $1.00 per million characters, and Cloudflare Workers AI and Groq are also in the open TTS price index with their sources and read dates.

Rate limits: 429 rate_limited with retry_after_seconds when exceeded.

Browser use

Text to speech API guides and comparisons

Next

Add speech to a web app · Industry agents · Your own voice as an API · Use your voice in your apps · Stream your first sentence

Questions, answered plainly

Which languages does text to speech support?

English, Spanish, French, Portuguese, Italian, Japanese and Chinese, with Hindi as a preview. Non-English clips are marked native-review-pending.

Can I use my own voice?

Yes. Record about 20 seconds or describe a voice and get a voice id; own-voice synthesis is $3 per million characters and currently English.

Do repeated sentences cost less?

A sentence we have already stored plays back in about 0.14 seconds. Offline packs go further: you buy them once and play them from your own device with no per-play fee.

Does it support speech marks for highlighting text?

Yes. Add marks=true to a request to get exact sentence times and word times for text generated live. The response says how the word times were made: model_aligned (about 15 ms from an independent aligner on average), aligned (measured from the audio, about 70 ms on average) or estimated (the fallback). They are not engine-exact like Amazon Polly's.