Model positioning
Qianfan-QI-VL is a vision large model for image quality inspection developed in-house by Baidu AI Cloud's Qianfan platform. It is available as a preconfigured service and is primarily intended for quality inspection of e-commerce product images. (qianfan.cloud.baidu.com)
Capabilities and uses
It supports detecting human body defects in AIGC, identifying mosaics and watermarks, and detecting trademarks. It can be used to inspect product images uploaded to e-commerce platforms. (qianfan.cloud.baidu.com)
Limitations
The official announcement focuses on e-commerce image quality inspection and does not specify accuracy, supported image formats and dimensions, context length, or a specific output format; performance in other scenarios has not been verified. Detection results should not be assumed to be error-free, and important results should also be reviewed by a human.
Release date
Baidu's official announcement states that the model was part of the Qianfan ModelBuilder update of June 13, 2025; the announcement page was published on June 16, 2025. (qianfan.cloud.baidu.com)
Pricing
As of November 18, 2026, no input or output pricing for this model had been verified on the Baidu domestic and international official pricing pages consulted. This does not establish that the service is free or has been discontinued. Domestic source consulted: Baidu Qianfan Model Service Billing; international source consulted: Baidu AI Cloud Model Pricing. (cloud.baidu.com)
Official source
Baidu AI Cloud Qianfan Community: Announcement of Qianfan Series Scenario Model Updates. (qianfan.cloud.baidu.com)