Mac-optimized YuE2 music generation using native MLX inference. Generate complete songs (melody, chords, vocals, accompaniment) from text prompts. Features an LLM Writing Room for lyric composition via OpenAI-compatible APIs (LM Studio, etc.). Pure Apple Silicon — no PyTorch needed.
Radar
New projects people are discovering or following across Pinokio.
Face swapping for images and video, with a React UI. Independent project; AGPL-3.0. Private — access is by invitation.
Steer a story in realtime, space pilot style, with Qwen3.5 2B running locally on llama.cpp. No API keys.
Text Generation model with 27B parameters. Download, explore files and run it on Hugging Face.
[NVIDIA ONLY] AllTalk-TTS is a unified UI for E5-TTS, XTTS, Vite TTS, Piper TTS, Parler TTS and RVC, based on CoquiTTS, including a finetune mode.
Customizing Realistic Human Photos via Stacked ID Embedding https://github.com/TencentARC/PhotoMaker
Silero Models: pre-trained text-to-speech models made embarrassingly simple
Fast, free voice assistant for your Mac. No API key. Remembers you and learns your shortcuts. Hold ⌃⌥Space and talk.
Launcher for Stable Audio 3 Small Music, Small SFX, and NVIDIA Medium using public cocktailpeanut Hugging Face mirrors. https://github.com/Stability-AI/stable-audio-3
Openvoice 2 Web UI - A local web UI for Openvoice2, a multilingual voice cloning TTS https://x.com/myshell_ai/status/1783161876052066793
Stable Diffusion WebUI Forge is a platform on top of Stable Diffusion WebUI (based on Gradio) to make development easier, optimize resource management, and speed up inference. https://github.com/lllyasviel/stable-diffusion-webui-forge?tab=readme-ov-file
Super Optimized Gradio UI for AI video creation for GPU poor machines (6GB+ VRAM). Supports Wan 2.1/2.2, Qwen, Hunyuan Video, LTX Video and Flux. https://github.com/deepbeepmeep/Wan2GP
An all-in-one, 100% local AI creative studio, director, and multi-track editor. Generate with MiniMax H3, LTX-2.5/2.3, Wan, Flux, Qwen, and more; turn an idea or song into a planned production; then finish it on the timeline. Requires an NVIDIA GPU (6GB+ VRAM).
[AMD ONLY] All-in-one local AI video, image & music studio built on the WanGP pipeline (Wan 2.1/2.2, LTX-2.3, Qwen, Hunyuan Video, Flux). ROCm-powered. Supported on Windows and Linux for RDNA 2-4 dGPUs (RX 6000/7000/8000/9000) and gfx1150-53 APUs (Strix Point/Halo, Krackan Point/Halo). 6 GB+ VRAM recommended. For NVIDIA GPUs use the upstream Maestro app. Credits: the app itself is Maestro by Blizaine (github.com/Blizaine/Maestro) — this is just an AMD installer wrapper around it. The clone-at-install-time approach used to port it to AMD was pioneered by 6Morpheus6's wan2gp-amd (github.com/6Morpheus6/wan2gp-amd).
Real-time face swap and one-click video deepfake with only a single image.
[AMD ONLY] All-in-one local AI video, image & music studio built on the WanGP pipeline (Wan 2.1/2.2, LTX-2.3, Qwen, Hunyuan Video, Flux). ROCm-powered. Supported on Windows and Linux for RDNA 2-4 dGPUs (RX 6000/7000/8000/9000) and gfx1150-53 APUs (Strix Point/Halo, Krackan Point/Halo). 6 GB+ VRAM recommended. For NVIDIA GPUs use the upstream Maestro app. Credits: the app itself is Maestro by Blizaine (github.com/Blizaine/Maestro) — this is just an AMD installer wrapper around it. The clone-at-install-time approach used to port it to AMD was pioneered by 6Morpheus6's wan2gp-amd (github.com/6Morpheus6/wan2gp-amd).
