Spex-TTS review
A developer-focused TTS API for conversational applications needing cloning and streaming.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
Spex-TTS is Gandr’s text-to-speech engine for developers and businesses building voice agents, narration, dubbing, live translation, and other audio applications. It is available through web and API interfaces, with REST, server-sent events, and WebSocket access. The service generates speech in 23 languages, supports voice cloning from about ten seconds of reference audio, and returns WAV or raw PCM audio. Its feature set is aimed at teams building conversational and real-time voice products rather than users seeking a basic, no-code narration tool.
Pricing ranges from a free plan with 50,000 tokens to start, to Take a Gandr at $10 per month for 1 million tokens and Pro at $50 per month for 5 million tokens. Both paid token plans reset monthly. Scale / flat rate costs $150 per stream per month with annual billing or $180 per stream per month with monthly billing, and provides one unlimited, unmetered stream. Enterprise includes volume, burst, and failover arrangements. The plan structure gives developers a low-cost entry point and a clear path to higher token allowances or an unlimited stream.
Spex-TTS stands out for real-time delivery and request-level control. Developers can use SSE or WebSocket streaming for conversational calls, while controls include temperature, CFG weight, seed, breaks, and spelling tags. Inaudible watermarking is applied to generated audio, and Gandr states that customer text and reference audio are not used for model training. Vapi is the listed integration. Teams that need formal certifications, published commercial-use permission, or a broader verified ecosystem should evaluate alternatives; teams building API-driven voice agents and expressive streaming workflows should consider Spex-TTS.
Spex-TTS pros and cons
- Where it wins
- Streaming through REST, SSE, and WebSocket interfaces
- Instant voice cloning from about ten seconds of reference audio
- 23 languages with per-request expression controls
- Where it doesn't
- Generated audio includes inaudible watermarking
- Formal certifications, including SOC 2, are not complete
- Published pages do not state explicit commercial-use permission
Spex-TTS fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update