Install size
Runtime disk footprint, excluding model weights.
Results as tablesRelease main-f03640150, as of 2026-11-13
A snapshot taken when this site was last built. The charts above read live data and may be newer. The same tables as markdown.
| Platform | Engine | Size | Version |
|---|---|---|---|
| Apple Silicon | llama.cpp | 31.3 MB | b11312 |
| Apple Silicon | LiteRT-LM | 38.7 MB | v0.19.1 |
| Apple Silicon | Pie | 103 MB | 0.5.4 |
| Apple Silicon | Splash | 171 MB | 1.1.0 |
| Apple Silicon | MLX | 359 MB | 0.31.3 |
| Apple Silicon | vLLM Metal | 1.9 GB | 0.33.0 |
| NVIDIA GPU | Pie | 1.2 GB | 0.6.0 (bb2d6bf43) |
| NVIDIA GPU | TensorRT-LLM | 12.3 GB | 1.4.0rc2.post1 |
| NVIDIA GPU | vLLM | 17.4 GB | 0.32.0 |
| NVIDIA GPU | SGLang | 19.8 GB | 0.5.22 |
| AMD GPU | Pie | 5.9 GB | 0.6.0 (5871e632) |
| AMD GPU | vLLM | 20.1 GB | 0.32.0 |
| AMD GPU | SGLang | 20.6 GB | 0.5.22 |
| Google TPU | Pie | 808 MB | 0.6.0 (bb2d6bf43) |
| Google TPU | SGLang | 2.2 GB | sglang-jax 0.1.0rc0 |
| Google TPU | vLLM | 4.9 GB | vllm-tpu 0.32.0 |
| Browser | llama.cpp WebGPU | 4.9 MB | b11312 |
| Browser | WebLLM | 14.5 MB | 0.2.87 |
| Browser | Pie | 35.2 MB | server-web 0.6.4 |