Coqui
Coqui was an open-source text-to-speech company from ex-Mozilla engineers, offering voice cloning and emotional speech synthesis through Coqui Studio.
RIP Coqui
Shut down · January 2024
Coqui was founded by engineers from Mozilla's shuttered speech team — Eren Gölge, Josh Meyer, Kelly Davis and Reuben Morais — and became the reference open-source text-to-speech stack, with XTTS handling voice cloning from seconds of audio across dozens of languages. The commercial Coqui Studio and its API wound down at the turn of 2024, with founders citing competition and a preference for open source over commercialisation. The models remain on GitHub and Hugging Face and are still widely used.
Last known pricing: Free tier plus paid Coqui Studio subscription (historical)
Is Coqui shut down?
Yes. Coqui shut down on January 3, 2024 and is no longer available. Its record is preserved in the ToolDirectory.AI graveyard, our hand-reviewed registry of AI tools that shut down or were acquired.
What to use instead
NaturalReader
NaturalReader is a text-to-speech tool that reads PDFs, documents, and webpages aloud with AI voices in 90+ languages.
Freemium
4.84
320
Okareo
Simulation testing for voice and text AI agents: synthetic users, audio-condition evals, CI/CD gating.
Free Trial
4.84
234
aiemployes
AIEmployes is a marketplace of ten preset AI agent workers — receptionist, support, sales, and admin roles — hired per agent on a subscription.
Paid - From $79/week per agent
4.77
380
Overview
What happened to Coqui
Coqui shut down its commercial business at the start of 2024. The paid Coqui Studio web app and the REST API were retired, with servers going offline in December 2023 and the closure confirmed publicly in January 2024.
Coqui was founded by engineers from Mozilla's discontinued speech team — Eren Gölge, Josh Meyer, Kelly Davis and Reuben Morais — and built what became the reference open-source text-to-speech stack. Its XTTS model cloned a voice from a few seconds of audio and spoke it across dozens of languages, and Coqui Studio wrapped that in a browser-based editor with emotion and style control aimed at game developers and dubbing.
Key Features
- XTTS voice cloning from a few seconds of reference audio
- Multilingual synthesis across dozens of languages
- Emotion and speaking-style control for character voice work
- Coqui Studio browser editor plus a REST API
- Apache-licensed open-source TTS toolkit on GitHub
Why it shut down
The founders pointed to financial pressure, intense competition in generative voice, and a decision to concentrate on open source and model development rather than commercialisation. Josh Meyer's note at the time observed that generative voice would only get better, faster and cheaper, and that open source would be a large part of that — an accurate forecast that also described the commodity pressure on Coqui's own paid product.
Unusually for a shutdown, the technology outlived the company. The open-source models stayed available on GitHub and Hugging Face, remain widely used in self-hosted voice pipelines, and continue to be forked and fine-tuned.
What to use instead
The original repository is still available for self-hosting, and community forks continue to maintain it. For a managed alternative, the voice AI and audio categories in our directory cover hosted text-to-speech and voice-cloning services with commercial licensing and support.