Qwen3.8-Max is the most capable flagship model in the Qwen family to date, with 2.4 trillion total parameters and 95 billion active parameters. Built on the Qwen 3.5 mixture-of-experts architecture, it targets long-horizon autonomous coding, scientific research, complex business automation, and agent workflows—and its weights are being released openly from day one.
Inside Tabbit, Qwen3.8-Max goes beyond answering questions. It can follow complex requirements, retain long context, and stay coherent across multiple rounds. That makes it a strong option for coding, reading papers, drafting plans, organizing research, brainstorming, and content creation.
Qwen3.8-Max is already available in Tabbit. Select “Switch model” in the lower-right corner of the chat box to try it.