01Generate
Text to Speech
Generate lifelike speech from text, with streaming, SSML, and emotion on the Simba models.
- Realtime streaming
- SSML and emotion
- Speech marks
SpeechifyAI Build
Every building block for speech. One API.
The developer API for voice AI: text to speech and voice cloning on the same speech stack that powers Speechify, exposed as building blocks you can compose.
500K characters free every month. Commercial use included, no card.
- Text to speech
- Voice cloning
- <100ms first byte
- One key, one bill
Synthesis request
Production REST endpoint
POST /v1/audio/speech
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"
}'- Model
- simba-3.2
- Voice
- sabrina
- Audio format
- mp3
Explore speech
Hear it before you write a line.
Hear prepared samples of expressive speech, voice cloning, and multilingual synthesis. Then open the playground to try your own text.
Simba 3.2 · Emotion control
Same words.
Different feeling.
Choose a delivery. Hear how the same voice changes its tone and rhythm.
Neutral.
Beatrice · Delivery sample
The quick brown fox jumped over the lazy dog.
Ready to listen
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
Building blocks
The complete speech layer, kept composable.
Generate a voice or create one. Both capabilities share one API, one key, and one bill, so the first prototype and the production integration stay on the same path.
02Create
Voice Cloning
Create a voice from a short sample, get a voice ID, and synthesize with it on the same endpoints.
- Consent verified
- One voice ID
- Same synthesis API
Benchmark
Independently evaluated, ready to build on.
The speech block runs on Simba 3.2, a production text-to-speech model evaluated on the independent, blind-listening Artificial Analysis leaderboard. The quality evaluation is independent, not self-reported.
- streaming text-to-speech
- Native
- per 1M additional characters on $10/month Starter
- $10
- self-serve languages
- 6
- included on paid plans
- Cloning
Model system
Choose for the experience, not the integration.
Both Simba models use the same endpoint. Select a model and a compatible voice without rebuilding the rest of your speech stack.
Recommended for English
Simba 3.2
Streaming-native speech for real-time playback.
Built for real-time playback, with finer-grained emotion control and SSML prosody, served from the full English voice roster with your cloned voices included. English only, and the recommendation for any English build.
- Delivery
- Streaming-native
- Playback
- Real-time
- Languages
- English
- Voice cloning
- Self-serve
Alternative
Simba 3.0
Multilingual streaming
Streaming-native across 6 languages and 7 locales. Set the language on simba-3.0 and one model covers the supported set, with zero-shot voice cloning.
- 6 languages
- 7 locales
- Streaming-native
- Voice cloning
Start building on Simba
Start with Simba 3.2 for English. Choose Simba 3.0 for its supported multilingual locale set.
Speech system
The details that turn audio into a product.
Streaming, identity, expression, synchronization, and output formats sit behind the same synthesis request. Add only the controls your experience needs.
Delivery
Streaming for realtime
Chunked audio that starts playing before the sentence finishes generating. Up to 20,000 characters a request.
Identity
Voice cloning
Create a voice from a short sample, get a voice ID, and synthesize with it on Simba 3.2 or Simba 3.0, on the same endpoints. Consent required.
Direction
SSML control
Full SSML (prosody, break, emphasis, sub) for exact control over delivery and timing.
Expression
13 emotion styles
Shift a single voice across narration, dialogue, and emotional registers with speechify:style.
Synchronization
Word-level speech marks
Timestamps for every word, returned alongside the audio, for synced captions and read-along highlighting.
Output
Five audio formats
wav, mp3, ogg, aac, and pcm, plus u-law from the stream endpoint for telephony.
Quickstart
From API key to production in three steps.
The production path is the quickstart path. Start with one request, then add realtime delivery and synchronization when the product calls for them.
- 01
Create one key
Open a workspace and start with 500K characters each month. Keep the key on your server.
SPEECHIFY_API_KEY
- 02
Send one request
Choose a model and voice, send text, and receive audio from the same endpoint you will use in production.
POST /v1/audio/speech
- 03
Move to production
Switch to streaming where latency matters, add speech marks when the interface needs timing, and keep the same key and billing model.
stream · observe · scale
Use cases
One speech layer, shaped around the experience.
Use the same models for a realtime reply, a long-form library, a recognizable brand voice, or localized audio. The surrounding product is yours to compose.
01
Conversational products
Stream a responsive voice into assistants, tutors, games, and interactive characters.
02
Narration at scale
Turn articles, lessons, books, and product content into consistent long-form audio.
03
A recognizable voice
Create a consent-verified voice identity and use it across every synthesized output.
04
Localized experiences
Ship spoken content across languages while keeping one integration and one account.
Pricing
One balance. The real unit on the bill.
Text to speech is billed per input character, with no credit conversion or token math. Your plan funds one balance across the speech capabilities you use.
- per 1M additional chars on $499/month Scale
- $6
- per 1M additional chars on $99/month Pro
- $8
- per 1M additional chars on $10/month Starter
- $10
- free every month
- 500K
Measured in input characters
No separate model surcharge
Commercial use on every plan
Start building free
No credit card required.
Enterprise
Built for teams at scale.
SOC 2 Type II, SSO, a signed DPA, and volume pricing, plus a Forward Deployed Engineer who builds alongside heavy-traffic and regulated teams. Most ship a first use case in two to three weeks.
FAQ
Common questions.
What is SpeechifyAI Build?
The developer API for voice AI, organized as building blocks. Text to speech and voice cloning are available today: generate lifelike speech, stream long-form audio, and clone a voice from a short sample, all on one API, one key, and one bill.
What can I build with it today?
Anything that needs synthesized speech: narration, agents, accessibility, captions, IVR, dubbing. Text to speech and voice cloning are live now on one key and one bill.
Which model should I use?
Use Simba 3.2 for English or Simba 3.0 for 6 languages across 7 locales. Both are streaming-native and support voice cloning on paid plans. Check each voice's model and language compatibility. Broader language coverage is available on request.
How much does it cost?
Paid plans have a monthly subscription and an included USD usage balance. Additional text-to-speech 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 the complete monthly costs.
Start building with voice.
Your first audio file is one request away. 500K characters free every month, commercial use included.
No credit card required.