Real‑time calls: ElevenLabs Conversational AI includes background noise removal (and echo cancellation) when you connect via WebRTC in the SDKs/embed. Set your client to use the WebRTC transport to benefit from these audio improvements. (elevenlabs.io)
Uploaded/recorded audio: ElevenLabs also offers a separate Voice Isolator tool and API to remove background noise and isolate speech from files (post‑processing). (elevenlabs.io)
If you need stronger “voice isolation” on existing recordings, use Voice Isolator; for live agents, use WebRTC for built‑in noise suppression during conversations. (elevenlabs.io)
Short answer: For PSTN phone calls (Twilio/SIP), ElevenLabs’ docs do not advertise built‑in background noise suppression/voice isolation. WebRTC sessions in the browser do include echo cancellation and background noise removal; that claim is specific to WebRTC, not telephony. For post‑processing or non‑real‑time cleanup, use Voice Isolator.
WebRTC path: “echo cancellation and background noise removal” is provided when using WebRTC with Conversational AI. (elevenlabs.io)
Telephony path: Phone integrations are documented via Twilio and SIP; docs cover formats (e.g., μ‑law 8 kHz) and setup, but do not mention built‑in denoise/voice isolation for live calls. (elevenlabs.io)
Offline/background cleanup: Voice Isolator (app/API) removes background noise from recorded audio, not live calls. (help.elevenlabs.io)
If you need noise suppression on a live phone call today, the official docs only confirm it for WebRTC browser sessions, not for PSTN telephony. (elevenlabs.io)
fading elk
quaint zenith