Qwen: Qwen3.5 397B A17B

Qwen3.5 397B A17B is a multimodal model from Qwen that accepts text, image, and video inputs, making it one of the broader input options available at this tier. It supports a 256,000-token context window, tool use, and reasoning, which covers the core requirements for agentic and long-document workflows. Structured output support is unconfirmed based on available data. At $0.385 per million input tokens and $2.45 per million output tokens, it sits in a mid-range price band. Its blended benchmark score of 78.3 across four benchmarks is modest overall, but the agentic subscore of 92.1 stands out and may interest teams building tool-calling or multi-step pipelines. Developers who prioritize video input support or long-context agentic tasks should shortlist it; those focused primarily on coding will find the coding subscore of 68.1 relatively unremarkable compared to specialized alternatives.