Model Positioning
A domain-specialized image understanding model from Baidu Qianfan, designed to recognize and extract information from a single image; the official community lists it as a model for the education domain. (cloud.baidu.com)
Capabilities and Uses
- Recognizes and extracts information from a single image, outputting it in YAML format for engineering processing.
- Can be used to extract image content in educational scenarios and for subsequent structured processing. (qianfan.cloud.baidu.com)
Limitations
- Official documentation focuses on single-image input; support for multiple-image input cannot be inferred from it.
- The official announcement does not specify the model's accuracy, supported languages, image size limits, or context length; current service availability is subject to the Baidu Qianfan platform. (cloud.baidu.com)
Release Date
On August 28, 2025, Baidu's model update log listed it as a newly added image understanding model for Inference Service API V2. (cloud.baidu.com)
Pricing
As of November 18, 2026, a search of Baidu AI Cloud's domestic model billing documentation and its international site pricing page did not verify an official price for this model; the billing methods, unit prices, and billing units for input and output were also not verified. The domestic and international pages are, respectively: Domestic model service billing and International site model pricing. Not being listed on the pages checked does not mean that an official price has never been published. (cloud.baidu.com)
Official Sources
- Baidu Qianfan model update log (cloud.baidu.com)
- Baidu Qianfan community product updates (qianfan.cloud.baidu.com)
- Direct access to the user-provided Baidu AI documentation URL timed out during this check; the Baidu AI Cloud update log above provided accessible information about the corresponding new model release.