SpeakSay
SpeakSay is a web text-to-speech service with an API for creating voiceovers and other audio projects.
What SpeakSay is
SpeakSay is a web text-to-speech service with an API for creating voiceovers and other audio projects. It offers more than 200 voices across more than 20 languages, plus voice cloning from an uploaded sample. SpeakSay says cloning can use a 30-second sample and recommends 60 seconds for best results, with V1 Classic, V2 Multilingual, and V3 Turbo models. The API can generate speech programmatically or stream it into apps, games, and demos. Speech controls include SSML, and downloads are available as MP3, WAV, or OGG. A no-sign-up demo allows 1,000 characters per day. SpeakSay says samples are encrypted, used only to create the user’s model, never shared, and deletable at any time; consent from the voice owner is required. The Free plan costs nothing and includes 10,000 credits, MP3 downloads, standard quality, basic voice selection, and community support. Paid plans include commercial licensing for generated audio.
Good
- More than 200 voices
- Voice cloning with stated sample guidance
- Developer API and streaming
- MP3, WAV, and OGG exports
- Free plan available
Watch out
- Free plan has basic voice selection
- Free plan includes MP3 downloads only
- Commercial license is stated for paid plans
Who it suits
SpeakSay suits creators making voiceovers and developers adding AI voice to apps, games, or demos. It also fits users who need voice cloning, multiple export formats, or programmatic generation.
Verdict
SpeakSay combines multilingual text-to-speech, voice cloning, speech controls, exports, and API access. Its free and paid plans provide different credit, support, and licensing terms for web and developer workflows.
Plans and prices
From SpeakSay's own pricing page, checked 9 Oct 2001.
| Plan | Price | What it gives you |
|---|---|---|
| Free | Free | 10,000 credits · Community support · MP3 download · Standard audio quality · Basic voice selection |
| Lite | $10 a month | 500,000 credits · 5,000 characters per generation · 5 cloned reference voice credits · 1 queued task · 1 month of conversion history · $10.00/month |
| Pro | $17 a month | 1,000,000 credits · 6,000 characters per regular conversion · 15 cloned reference voice credits · 3 queued tasks · Priority 36h customer support · $17.00/month |
| Studio | $24 a month | 3,000,000 credits · 7,000 characters per regular generation · 30 cloned reference voice credits · 4 queued tasks · Priority 24h email support · $24.00/month |
| Lifetime | $150, paid once | 2,000,000 characters per month · 50 cloned reference voice credits · 50 Remix Voices · 4 queued tasks · $150.00/once |
The facts
Each one read from SpeakSay's own pages, with the date it was checked.
What it does
- Audio formats
- The English text-to-speech page says users can download audio as MP3, WAV, or OGG.checked 9 Oct 2001
- Cloning models
- The page describes V1 Classic, V2 Multilingual, and V3 Turbo voice-cloning models.checked 9 Oct 2001
- Commercial use
- SpeakSay says audio generated on paid plans includes a full commercial license.checked 9 Oct 2001
- Developer API
- SpeakSay describes an API for generating speech programmatically and streaming it into products.checked 9 Oct 2001
- Support
- The pricing page lists community support for Free, priority 36-hour customer support for Pro, and priority 24-hour email support for Studio.checked 9 Oct 2001
- Text to speech
- SpeakSay turns text into natural-sounding speech for voiceovers and other audio projects.checked 9 Oct 2001
- Voice cloning
- Users can upload a voice sample and generate speech in the cloned voice.checked 9 Oct 2001
- Voice consent
- SpeakSay says users must have the voice owner's consent and its terms prohibit impersonation or unconsented cloning.checked 9 Oct 2001
- Voice data
- SpeakSay says voice samples are encrypted, used only to create the user's model, never shared, and the clone can be deleted at any time.checked 9 Oct 2001
- Voice library
- SpeakSay says it offers more than 200 voices across more than 20 languages.checked 9 Oct 2001
- Voice sample requirement
- SpeakSay says voice cloning can use a 30-second sample and recommends 60 seconds for best results.checked 9 Oct 2001
- Who it serves
- SpeakSay presents its developer offering for adding AI voice to apps, games, and demos.checked 9 Oct 2001
Money
- Free demo
- The English text-to-speech page says users can generate 1,000 characters per day without signing up.checked 9 Oct 2001
The company
- Headquarters
- Ashton-under-Lyne, United Kingdomchecked 28 Sep 2001
Alternatives to SpeakSay
- 1
Resemble AI
Free planAndroidiOSWebAPI8.3From $280 a month - 2
Adobe Firefly
Free planAndroidiOSWebAPI7.7Free plan - 3
Google Cloud Text-to-Speech
AndroidiOSWebAPI7.7From $4 a month - 4
ElevenLabs
Free planWebAPI7.6From $6 a month - 5
Neocortex
Free planAndroidiOSWebAPI7.4From €9 a month - 6
Altered Studio
Free planWebAPImacOSWindows7.3From $30 a month