Qwen: Qwen3.5-27B

Qwen3.5-27B is a multimodal model from Qwen that accepts text, image, and video inputs with a 262,144-token context window. It supports tool use and reasoning, making it usable for agentic workflows and multi-step tasks. Structured output support is unconfirmed based on available data, so developers who depend on that feature should verify before committing. At $0.195 per million input tokens and $1.56 per million output tokens, it sits in the budget-to-mid range for capable models. Its blended benchmark score of 70.5 across only 3 benchmarks limits confidence in that figure, though its agentic score of 90.1 is notably high. Teams running tool-heavy or agent-based pipelines on a tighter budget have a reasonable case to shortlist it, while those needing broad, well-documented benchmark coverage may want to wait for more data.