Shypd.aiThe largest open index of AI toolsAnubis OSS is an open-source tool for local LLM benchmarking on Apple Silicon. It provides real-time hardware telemetry, correlating GPU, CPU, power, and memory metrics with inference performance.
Trending
About
Anubis OSS is a native macOS application designed for comprehensive benchmarking of local Large Language Models (LLMs) specifically on Apple Silicon. It offers real-time hardware telemetry, correlating GPU, CPU, power, memory, and frequency data directly with LLM inference performance. The tool features a detailed benchmark dashboard with over 8 metric cards, 30 data points, and 11 live charts, including power telemetry and configurable prompt presets. Users can export and compare results, stream responses with live hardware overlays, and utilize an Arena Mode for side-by-side A/B model comparisons. Anubis also includes a Vault for unified model management across various backends and provides detailed power telemetry for GPU, CPU, ANE, and DRAM, allowing for watts-per-token efficiency analysis. It supports various OpenAI-compatible inference endpoints like Ollama, oMLX, LM Studio, and vLLM.
Capabilities
Pricing & Plans
Open Source
Open Source
FAQs
Trending
