Magnitude
Open source inference server that runs local AI models for your hardware, integrated with agents like Cline.
0 members list Magnitude, unchanged over the last 12 weeks, data as of 18 November 2026
Members who list Magnitude in their Stack, by week.
Members who list Magnitude
Be the first to add Magnitude
Nobody on Stackness runs Magnitude yet. Add it to your stack and you're the first member running it.
About Magnitude
Magnitude is an open source inference engine designed to run large language models locally on your hardware. It optimizes itself for your specific device by compiling and tuning computation kernels on your machine before models run, allowing open source models to execute efficiently. The tool works with various hardware setups including Apple Silicon, NVIDIA and AMD GPUs, or CPU-only systems. Magnitude integrates with agent applications like Cline through a simple connection process and provides an OpenAI-compatible API for broader compatibility. The inference engine emphasizes privacy, keeping prompts, files and models entirely on your machine with no internet requirement once a model is downloaded.
Developed by magnitudedev, Magnitude is licensed under Apache 2.0 and is available as a desktop application. The tool supports a range of popular open-weight model families with hand-optimized kernels for faster inference. It manages memory efficiently, automatically freeing resources when agents stop, and enables multiple concurrent sessions that share cached prefixes to maintain performance. The project includes comprehensive documentation and offers downloads for macOS, Linux and Windows platforms.
Checked 16 November 2026.
Nobody on Stackness runs Magnitude yet.
Sign up and add itPosts about Magnitude
- Mistral Large 4 vs Beam vs Kolibri: which open weights you can download, and what your hardware can serve
Only Kolibri's weights are out. Beam and Mistral Large 4 stay API previews until late October, and Kolibri's 4-bit build runs at 50 tok/s on a 64GB Mac.
- Llama.cpp alternatives for running local agents: ds4, Magnitude and Strata against llama.cpp, vLLM and Ollama
ds4 runs big MoE models on 96 GB Macs, Magnitude tunes kernels to your device, Strata puts one model on a gaming PC. Every speed claim is self-reported.