NVIDIA/personaplex

This page introduces an external open-source repository. It is not an HDATF product.

What it does

PersonaPlex is a real-time, full-duplex speech-to-speech conversation model. A text role prompt sets the persona and an audio sample conditions the voice. It is based on the Moshi architecture and weights.

How it helps ATF

Not related to our products. Listed because it appeared in the ranked lists.

License

MIT Permissive. Commercial use and changes are allowed if the copyright notice is kept.

More in this category

• QwenLM/Qwen3-TTS
  An open-source series of text-to-speech models from the Qwen team at Alibaba Cloud. It supports streaming speech generation, free-form voice design, voice cloning and voice control through natural language.
• remotion-dev/remotion
  A framework for making videos with React code, where the code is the source of truth. Videos can be made with a coding agent, edited by drag and drop, or tied to data and batch-rendered on your own infrastructure.
• ace-step/ACE-Step-1.5
  ACE-Step 1.5 is an open-source music generation model that runs locally on Mac, AMD, Intel and CUDA devices. Users can train a LoRA from a few songs to capture their own style.
• torvalds/AudioNoise
  AudioNoise is a set of simple digital audio effects, such as phaser and echo, written to learn the basics of audio processing. It uses IIR filters and delay loops that take one sample in and give one out.
• dwzhu-pku/PaperBanana
  A tool that generates illustrations for academic papers with a vision-language model and an image generation model, offering a Streamlit interface for choosing models. It is a fork of Google Research's PaperVizAgent.

Only repositories in the ranked Trendshift lists are included, and the lists are used only to find candidates. We do not copy their ranks. Descriptions, licenses and star counts come from each GitHub repository. The notes are our own reading. We have not tested these projects, and a place on a trending list does not prove quality.