Murf AI is a voice platform with two audiences: creators, who use its studio and apps to produce voiceovers and dubbed videos, and developers, who call its text-to-speech API. The API has HTTP streaming and WebSocket modes and serves 150+ voices in 35 languages with 20 or more speaking styles, and a multilingual technology lets one voice move between languages while keeping native pronunciation.
Two speech models are offered. Falcon 2 is a streaming model built for conversational AI and real-time voice agents, with latency reported below 100 ms. Gen2 targets studio-quality synthesis with deeper customization. Related APIs cover a voice changer, dubbing automation, and translation, and Murf also sells a hosted voice-agent product for use cases such as receptionists, recruiters, and call centers.
Features
- Text-to-speech API: HTTP streaming and WebSocket delivery
- Falcon 2: a streaming model for real-time agents with sub-100 ms latency
- Gen2: a studio-quality model with advanced voice customization
- Voice library: 150+ voices in 35 languages with 20+ speaking styles
- Voice controls: pitch, rate, and style parameters per request
- Multilingual voices: a single voice can switch language with native pronunciation
- Voice changer, dubbing, and translation APIs: transform recordings and localize video and audio into 20+ languages
- Voice cloning: clone a voice for use in voiceovers
- SDK and integrations: an official Python SDK, an MCP server, and integrations with Pipecat, LiveKit, Agora, n8n, Make, and Zapier
- Creator apps: Canva, Google Slides, PowerPoint, and Adobe Captivate integrations
- Enterprise deployment: on-premise option and published rate limits
