KugelAudio

KugelAudio

Real-time text-to-speech model you can self-host

224 followers

Visit website
•
Most natural real-time TTS with voice cloning and sub-60ms latency, on-prem or via API. Grammar-aware normalization reads phone numbers, IBANs, addresses, and medications naturally across 25+ languages, with word-level timestamps and IPA support. Adapters for LiveKit, Pipecat, and Vapi. Built by 4 in Berlin.

KugelAudio

Launch date
Kugel Ear
Kugel EarSpeech-to-text that only asks when it's unsure

Launched on October 22nd, 2026

Kugel 3.5
Kugel 3.5Text-to-speech that beats any other Model out there

Launched on October 14th, 2026

Clarity
ClarityMute the room, keep the speaker, in real time

Launched on October 1st, 2026

KugelAudio
KugelAudioReal-time text-to-speech model you can self-host

Launched on May 28th, 2026