Video voiceover
Narration for explainers, ads, social clips and product demos from a script, with a tone that matches the piece and no studio booking. Media teams turn articles and posts into audio at newsroom speed.
SpeechifyAI · Text to speechText-to-speech API.
Text-to-speech API.
Made to feel spoken.
Expressive speech for your application, with streaming-native models, clear usage pricing, and one API for catalog and cloned voices.
500K free characters to start. Commercial use, no credit card.
bash
curl -X POST https://api.speechify.ai/v1/audio/speech \
-H "Authorization: Bearer $SPEECHIFY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"input": "Ship production voice in a few lines of code.",
"model": "simba-3.2",
"voice_id": "sabrina",
"audio_format": "mp3"
}'Example outputSpeech audio
MP3 in JSON- Independently evaluated on the Artificial Analysis TTS leaderboard
- 56 ms p50 first byte on Simba 3.2, production US East, measured 15 Sep 2026
- POST /v1/audio/speech and POST /v1/audio/stream
- 5 audio formats, SSML pauses and speaking rate
- Additional usage from $6 per 1M characters on Scale; monthly plan required
- Simba 3.2: English. Simba 3.0: 6 languages, 7 locales. Both streaming-native with paid voice cloning
Listening labVoices you can hear.
Voices you can hear.
Before you integrate.
Hear prepared Simba samples across language and voice identity. Compare the recordings, then take your own text to the playground.
Simba 3.0 · Voice cloning
A new script.
Your voice.
Start with a consented reference recording. Create a reusable voice for the words that come next.
Explore voice cloning01 / Reference
Original speaker
Ready to listen0:00
02 / Clone
Cloned voice
Ready to listen0:00
Two different recordings. Generated sample: Simba 3.0. Only clone a voice you have permission to use.
Simba 3.0 · Multilingual synthesis
Make yourself
understood.
Hear 6 language samples, each with a voice from its own catalog. These are separate samples, not translations of one recording.
Explore language supportEnglish
en-US
German
de-DE
Spanish
es-MX
French
fr-FR
Italian
it-IT
Portuguese
pt-BR
Prepared Speechify samples. Choose a sample, then press play.Try your own text
Build with these voicesFree tier, no credit card.
The signal
A speech system measured in more than speed.
Listening quality, latency, control, and cost belong in the same production decision.
Two independent benchmarks time Simba 3.2 from their own runners to the first sound: Coval measured a 106 ms median time to first audio over the 24 hours to 24 Sep 2026, and Voice Arena read 123 ms p50 on its US English board the same day. Their readings include the network to each runner and any silence before the first sound, so they read higher than our own first byte, and neither is its source. Our reviewed snapshots: Coval, Voice Arena.
- on Artificial Analysis's blind-listening leaderboard
- Evaluated
- per 1M additional characters on Scale
- $6
- self-serve languages, voice compatibility varies
- 6
- streaming first byte
- <100ms
Why developers choose Speechify
Top-tier voice, without the top-tier bill.
Production-quality voice
Simba 3.2 is independently evaluated on Artificial Analysis's blind-listening leaderboard. The quality claim is tested, not self-reported.
Clear usage pricing
Paid plans combine a monthly subscription, an included USD usage balance, and a published per-character rate for additional synthesis.
Drop-in compatible
One REST API with Python and TypeScript SDKs. Move off your current provider with minimal code changes.
Streaming-native
Built for real-time audio, with SSML pauses and speaking rate, and zero-shot voice cloning on paid plans.
The math
Compare speech usage at your volume.
Compare character volume at the listed usage rates. This is a usage-value estimate, not a monthly bill: subscription fees, included balances, discounts, and taxes are excluded. Your plan determines the rate and balance available to you.
Your monthly volume
10million characters / mo
≈ 1.7 million words
1M chars200M chars
Compare against
SpeechifyAI · usage valuePAYG
$100/mo
$10/1M PAYG · from $6/1M on Scale
- Flat per-character billing
- Voice cloning on paid plans
- Commercial use on every plan
- Characters, not tokens
Usage only. Your subscription fee and included balance are not applied here.
ElevenLabs · Flash v2.5 on the API
$400/mo
API rates read 28 Sep 2026: Flash v2.5 $40 per 1M, Eleven v3 $80
4×
the SpeechifyAI rate per character
Rates vary by plan tier, model, and annual commitment.
SpeechifyAI is shown at $10/1M characters on pay-as-you-go. The Scale usage rate starts at $6 per 1M with a paid subscription. Competitor rates reflect the listed model or tier, not every available plan. See plan fees and included balances.
Usage value at common volumes
| Monthly volume | SpeechifyAI ($10/1M) | ElevenLabs ($40/1M) | Rate difference |
|---|---|---|---|
| 1M chars | $10/mo | ~$40/mo | $30/mo$360/yr |
| 10M chars | $100/mo | ~$400/mo | $300/mo$3,600/yr |
| 100M chars | $1,000/mo | ~$4,000/mo | $3,000/mo$36,000/yr |
Start building
From key to production in three steps.
Get a key
Sign up and issue a key in under a minute. No credit card, and the free tier is cleared for commercial use.
Make the first call
One REST endpoint, with Python and TypeScript SDKs. The sample above runs as-is once your key is in.
Go to production
Rates drop as your volume grows, and SOC 2 Type II, SSO and a DPA are ready when procurement asks.
Use casesOne speech layer.
One speech layer.
Many products.
The endpoint stays the same. The voice, timing, format, and delivery adapt to the experience around it.
Podcasts
A written script becomes a produced episode: assign a voice per speaker, synthesize the lines, assemble the audio. Solo shows, co-host banter and interview formats without coordinating a recording.
Audiobooks and publishing
A manuscript becomes a finished audiobook with one consistent voice across every chapter, and a publisher's whole catalog runs through the same pipeline instead of a studio session per title.
Accessibility and read-aloud
Articles, documents, app screens and web pages spoken aloud with a follow-along cursor from speech marks, for readers with dyslexia, low vision or a preference for audio.
E-learning and training
One consistent narrator for every module, captions from speech marks, and multilingual synthesis with Simba 3.0 across 6 languages. Check voice and locale compatibility before localizing. A revision is a text edit and a re-render, not a re-recording.
Games
Thousands of lines and dozens of characters voiced from a dialogue script without a casting session, and lines generated at runtime for scenes that are not fixed at ship time.
IVR and phone prompts
Menu options, hold messages, announcements and after-hours notices generated from text in the format a phone system expects, updated in seconds rather than studio sessions.
Spoken notifications
A delivery update, an appointment reminder, a system warning or an in-app cue becomes a clear spoken message, synthesized from whatever the alert says rather than a fixed library of recordings.
Enterprise
Enterprise-ready when you are.
Start on the free tier today and bring procurement in later. SOC 2 Type II, SSO, a signed DPA and volume pricing are all here when you need them, and teams with heavy traffic or a regulated surface get a Forward Deployed Engineer building alongside them. Most ship their first use case in two to three weeks.
FAQ
Frequently asked questions
What is a text-to-speech API?
A text-to-speech API turns written text into spoken audio over a network request. You send text and a voice choice; it returns audio, either as a complete file or as a stream. SpeechifyAI's API adds SSML control, voice cloning, and speech marks, on a model evaluated on independent blind-listening leaderboards.
How fast is it?
Simba 3.2 returns its first byte in 56 ms at the median and 102 ms at the 90th percentile, measured on the production US East streaming path on 15 Sep 2026; the site rounds that to <100ms. It is streaming-native, so audio starts playing before the full clip is generated, which is what real-time apps like voice agents and live captioning need.
Has anyone measured it independently?
Yes. Two independent benchmarks time Simba 3.2 from their own runners to the first audible sound: Coval measured a 106 ms median time to first audio over the 24 hours to 24 Sep 2026, and Voice Arena read 123 ms p50 on its US English board on 24 Sep 2026. Their readings include the network to each runner and any leading silence, so they sit above our own first byte, which they do not source.
Can it control pacing and pauses?
Yes. Standard SSML directs the timing word by word: break for pauses, prosody rate for speaking rate, and sub for pronunciations, at no extra cost. Simba 3.2 and Simba 3.0 accept pitch, volume, emphasis and the speechify:style emotion tag but do not apply them.
What does it cost?
Paid plans have a monthly subscription and an included USD usage balance. Additional usage is $10 per 1M characters on Starter, $8 on Pro, and $6 on Scale. The Free plan includes 500K characters a month with commercial use. Voice cloning requires a paid plan. See Pricing for subscriptions, included balances, and additional usage.
Which models and languages can I use?
Simba 3.2 supports English. Simba 3.0 supports 6 languages across 7 locales. Both models are streaming-native and support voice cloning on paid plans. A catalog voice is not necessarily compatible with every model and language; check the voice metadata before synthesizing.
Which audio formats are supported?
Five: wav, mp3, ogg, aac, and pcm. The batch endpoint returns a full file including WAV; the streaming endpoint returns chunked mp3, ogg, aac, pcm, or u-law for telephony.
Start building in minutes.
Issue a key, paste the sample, and hear your first request come back. Free while you build, and priced per character the day you ship.
500K free characters to start. Commercial use, no credit card. SOC 2 Type II.