Tip Of AI
Discover what you can do with AI
RigSpark is a free and open source local AI platform that helps users discover which language, image, video and voice models their computer can run. It detects available RAM, GPU resources, disk capacity and Apple Silicon capabilities, then ranks models using hardware aware calculations. Each model receives a yes, slow or no verdict with estimated memory requirements and token speed ranges. The platform supports text chat, coding, writing, image generation, short video generation, spoken responses and voice dictation. It works with five runtimes including Ollama, llama.cpp, MLX, BitNet and LM Studio. Images and videos run through a local ComfyUI installation, while voice features use Kokoro, Chatterbox and Whisper. The catalog contains 254 models and also supports GGUF models from Hugging Face. RigSpark downloads and verifies model files using SHA 256 checksums, starts services on the local loopback address and supports offline recommendations. It includes context aware memory sizing, KV cache calculations, portable chat memory, local agents, skills, MCP connectors, terminal streaming and a browser workspace. The community edition is available for macOS and Windows, with command line support for Linux and additional installation options through Homebrew, Cargo and Docker. It is built in Rust and distributed under the MIT license.