Qwen: Qwen3 Next 80B A3B Thinking

Qwen3 Next 80B A3B Thinking is a text-in, text-out model from Qwen with a 262,144-token context window and a 32,768-token output ceiling. It supports tool use and reasoning natively, making it suited for multi-step tasks, agentic workflows, and long-document processing. Structured output support is unconfirmed, so developers who depend on strict schema compliance should verify that before committing. On the comparison front, this model has no independent benchmark coverage yet, so its relative quality against peers is unverified. Pricing is modest at $0.0975 per million input tokens and $0.78 per million output tokens, which positions it as a low-cost option for teams willing to run their own evaluations. Buyers who want third-party validated performance should wait for coverage to emerge; those comfortable testing internally may find the price-to-context ratio worth exploring.