Public APIs

Text to speech API for voice agents. One voice in 23 languages, every render watermarked.

Gandr's website screenshot

About Gandr

Gandr is a text-to-speech and voice cloning API that streams audio over WebSocket for voice agents (integrating with platforms such as LiveKit, Pipecat, Vapi, Retell, and Daily) and over HTTP for narration and dubbing use cases. Its underlying engine, Spex-TTS, supports 23 languages and produces a voice clone from roughly ten seconds of reference audio, without a separate training job. The service reports a first-audio latency of 116 ms p50 (server side) and 146 ms over the open internet.

Pricing is structured as a flat per-stream rate rather than per-character usage, so cost does not scale with volume of speech generated on a stream. Every generated clip carries an inaudible watermark, and submitted audio is not used for training. The infrastructure is described as SOC 2 Type 2, HIPAA-, and GDPR-verified, with per-key access quotas and revocation, and watermarking positioned as meeting EU AI Act Article 50 transparency requirements.

The product page lists agencies, startups, developers, enterprises, and game studios as target users, and cites use cases including AI customer support, AI receptionist, live translation, video dubbing, audiobook narration, and e-learning narration. A free tier is available to start, with custom pricing for accounts exceeding 500 lines arranged by call.

Key features

  • Text-to-speech and voice cloning over WebSocket and HTTP for platforms like LiveKit, Pipecat, Vapi, Retell, and Daily
  • Voice cloning from about a 10-second reference clip, no training job required
  • Streaming speech in 23 languages
  • 116 ms p50 first audio latency, server side warm
  • Unlimited/unmetered usage per stream (flat per-stream pricing, not per-character)
  • Inaudible watermark stamped into every generated clip

Frequently asked questions

Is there a free tier?

Yes, users can start free and sign in to get an API key issued on the spot.

How is Gandr priced?

Pricing is per stream rather than per usage, so cost stays flat regardless of volume; above 500 lines pricing is arranged on a call.

What platforms and integrations does it support?

It integrates with LiveKit, Pipecat, Vapi, Retell, and Daily, and works over WebSocket for agents or HTTP for narration and dubbing.

What languages are supported?

Gandr supports text-to-speech and voice cloning in 23 languages, including English, Spanish, French, German, Portuguese, Arabic, Chinese, and Japanese.

Is the service secure and compliant?

It runs on SOC 2 Type 2 audited infrastructure that is HIPAA- and GDPR-verified, with watermarked audio for EU AI Act Article 50 transparency and per-key quotas.

Ad
SerpApi - Search API favicon

SerpApi - Search API

SerpApi's Search API makes it easy and fast to scrape Google and other search engines.

Learn more

Featured products

  • SerpApi - Search API favicon

Show your product to thousands of developers

· 100k monthly pageviews
· 7k newsletter subscribers

Advertise your product