Ternary Bonsai WebGPU
🌳
Run 1.58-bit Bonsai LLMs locally in your browser on WebGPU
Transformers.js V4 demos
A collection of demos built with Transformers.js V4
-
-
Gemma 4 WebGPU
🚀4Run Gemma 4 locally in-browser on WebGPU w/ Transformers.js
-
Bonsai Image WebGPU
🌳State-of-the-art image generation, in your browser.
-
Privacy Filter WebGPU
🕵PII detection and text masking in your browser
-
Qwen3.5 0.8B WebGPU
🦫Describe what the camera sees in a few sentences
-
Qwen3.5 WebGPU
😻Run Qwen3.5 (0.8B, 2B, 4B) in-browser with Transformers.js
-
Qwen3.5-4B clawd-rift · Pi coding agent
🦾Pi coding-agent driven by clawd-rift Q4_K_M
-
Pi Agent · Soyuz (llama.cpp WebGPU)
🛰In-browser coding agent — Soyuz 4B via llama.cpp WebGPU
-
ml-intern MoE-100M (Ultra-FineWeb) — demo
🌊99M-active MoE on Ultra-FineWeb, mid-100B run
-
MoE-200M (Qwen3) — WebGPU demo
🌊620M MoE (200M active) — browser inference on WebGPU
-
MoE-200M Path B WebGPU Demo
🛣Generate text using a 200M MoE model in your browser
-
My Pi Agent
🦀29Generate full HTML web apps from text prompts