Base44 Superagents-AI agent that does it all
Mini toolsQwen3.5 4B
By Alibaba
Qwen3.5-4B is a post-trained vision-language model with 4B parameters, 32 layers, a hybrid Gated DeltaNet plus Gated Attention layout, and multi-token prediction training, with 262,144 native context length and stated extensibility up to about 1,010,000 tokens.
New
Multimodal
Gen 3
Overview
Qwen3.5-4B is a mid-size vision-language causal model with a vision encoder, designed for multimodal reasoning, coding, and agent workflows with very long context.
About Alibaba
Chinese e-commerce and cloud leader behind Taobao, Tmall, and Alipay.
View Company ProfileTools using Qwen3.5 4B
No tools found for this model yet.
Last updated: March 2, 2026
