Qwen 3.5 27B
Apache 2.0Alibaba · 27B · transformer-decoder
2026-03-15262K context27B params
Use Cases
chatcodereasoningmultilingualvisiontoolsmathwritingsummary
Quantization Options
About this model
Qwen 3.5 27B is a powerful dense multimodal model that excels across text, vision, coding, and agentic tasks. With 256K context and 201-language support, it's one of the most versatile models in its size class.
At Q4 it needs about 19 GB VRAM, fitting well on a RTX 3090/4090 or a Mac with 24 GB+ unified memory.