model
Bonsai 27B WebGPU Kernels
Running a 27B LLM in-browser is impressive but limited by WebGPU support and 1-bit quantization, suitable for experimental privacy-focused use.
6.6Overall
Utility7
Onboarding7
Craft6
Niche fit8
Longevity5
Five dimensions scored independently; overall is a weighted average. Scores are only comparable within this same rubric.
Good for
Developers and researchers needing offline, privacy‑first LLM inference.
Not for
Enterprise use cases requiring high accuracy and production‑grade performance.
Project description
Run a 1-bit 27B LLM locally in your browser on WebGPU
Alternatives
llama.cppWebLLMTensorFlow.jsONNX Runtime Web