Overview
Qwen Vision-Language models that support images, videos, and audio inputs.
Class Signature
from qwen_agent.llm import get_chat_model
llm = get_chat_model({
'model': 'qwen-vl-max',
'model_type': 'qwenvl_dashscope',
'api_key': 'your-api-key'
})
Configuration
Model: ‘qwen-vl-max’, ‘qwen-vl-plus’, ‘qwen-vl-ocr’
Must be ‘qwenvl_dashscope’
Usage Example
from qwen_agent.llm import get_chat_model
from qwen_agent.llm.schema import Message, ContentItem
llm = get_chat_model({
'model': 'qwen-vl-max',
'model_type': 'qwenvl_dashscope',
'api_key': 'your-api-key'
})
# Image input
messages = [
Message(role='user', content=[
ContentItem(text='What is in this image?'),
ContentItem(image='https://example.com/image.jpg')
])
]
for response in llm.chat(messages=messages):
print(response[-1].content)
- Images: URLs or local file paths
- Videos: Video file URLs
- Audio: Audio files (select models)
See Also
