Ternary bonsai 27b gguf local LLM inference using optimized 1.58bit weights. Derived from Qwen3.6-27B, it supports multi-step reasoning, agentic workflows, and a 262K context window directly on consumer hardware. Download direct Hugging Face repository model configuration files.
bonsai ternary-computer opendesktop qwen-api ollama-tools qwen-3b qwen-ai qwen-tts qwen-image-edit qwen-image-edit-2511 ollama-llama-3 qwen-tts-realtime qwen-endpoint qwen-models ternary-language qwen-workflow qwen-2-api bonsai-27b bonsai-27b-ollama bonsai-27b-gguf
-
Updated
Jul 18, 2026 - TypeScript