Qwen: Qwen3.6 27B

Qwen3.6 27B is a multimodal model from Qwen that accepts text, image, and video inputs, making it one of the broader input options at this tier. It supports tool use and reasoning, and its 262K-token context window accommodates long documents or extended multi-turn sessions without truncation. Structured output support is unconfirmed from available data. On benchmarks, the model scores 65.6 across just three benchmarks, so treat that figure as preliminary rather than a settled verdict. Its coding score of 88.6 is a meaningful signal for developer workloads, while its agentic score of 44.6 is notably lower. At $0.29 per million input tokens and $3.17 per million output tokens, the output cost is moderate but worth tracking on verbose tasks. Teams doing image or video analysis alongside code generation have reasonable grounds to shortlist it; those prioritizing agentic pipelines should compare alternatives with stronger coverage in that category.