Qwen: Qwen3.8 27B

Qwen3.8 27B is a model from Qwen that accepts text, image, and video input and produces text output. It offers a 1,000,000-token context window and allows up to 235,929 completion tokens per response, which suits long documents, large codebases, or extended video transcripts. It supports tool calling and a reasoning mode; structured output support is not listed, so verify that before relying on schema-constrained responses. The pricing is lopsided: input costs $0.04 per million tokens while output costs $2.30 per million. That makes it most economical for workloads that read far more than they write, such as retrieval over large contexts, summarization, or multimodal analysis with short answers. Output-heavy uses like long-form generation or verbose reasoning will cost noticeably more. Its blended benchmark score of 55.6 rests on a single independent benchmark, so its standing is only lightly evidenced. Shortlist it for input-heavy multimodal tasks and test it against your own workload before committing.