Qwen vision-language model for visual reasoning, documents, and agent tasks

Details

FieldValue
idQwen/Qwen2.5-VL-72B-Instruct
modalitytext+image->text
context_length128000
max_input_tokens120000
max_output_tokens8192
quantizationfp8
knowledge_cutoff2024-12
release_date2025-01-20
is_open_weightsyes
supports_tool_callyes
supports_reasoningno
supports_structured_outputyes
supports_attachmentyes
input_modalitiestext, image
output_modalitiestext

Providers