Qwen3-VL-235B-A22B-Instruct: Ultra-large vision-language, strong reasoning
Qwen3-VL-235B-A22B-Instruct is part of the Qwen lineup from Qwen, designed for production workloads that need reliable quality at scale. Ultra-large vision-language, strong reasoning.
It accepts Image, Text, Video inputs and produces Text outputs, with a context window of up to 128,000 input / 8,192 output tokens. The model is tuned for chat, analysis, and tool use in enterprise pipelines.
Native vision understanding supports charts, screenshots, UI mockups, and short video clips alongside text prompts.
