PersonaPlex-7B-MLX-4bit is a 7B parameter speech-to-speech model based on Moshi, optimized in 4-bit for MLX and Apple Silicon. It enables full-duplex conversational audio interactions and is distributed on Hugging Face for local inference. The model includes support for English and comes with tools for integration into custom voice applications.
In the Voice, TTS & speech space, PersonaPlex 7B takes a focused approach. It focuses on running real-time voice-based AI personas locally on Apple devices without cloud dependency. It is built as an open-source project for AI developers and researchers. PersonaPlex 7B is open source under the Apache-2.0 license. PersonaPlex 7B is available on the web, the command line, and API.
It is developed by aufklarer, and the product first shipped in 2026. Development happens publicly on GitHub with 1.1k stars and 427 commits in the last 90 days. Key capabilities include Full-duplex Conversation, speech-to-Speech, and MLX Optimization.
Latest indexed changes and source events
aufklarer/PersonaPlex-7B-MLX-4bit verified by the PulseGate indexer
Other apps tracked under the same category.