Model Positioning and Capabilities
Qianfan-Agent-Speed-32K is an agent-specific model developed in-house by Baidu. It is instruction-tuned for enterprise large-model applications and supports a 32K context. Baidu officially positions it for question-answering and agent scenarios, including knowledge question-answering, dialogue, and agent workflows. (ai.baidu.com)
Release Date
2025-01-02: Recorded as the release date of this version according to Baidu's official documentation, not as an independently verified first release date. (ai.baidu.com)
Documentation note: Based on the current data, whether this model was first released on 2025-01-02 has not been independently verified. The date is based on the description of this version in Baidu's model list. (ai.baidu.com)
Limitations
The context length is 32K; refer to the official documentation and platform for specific API parameters and availability. Model responses may be inaccurate, and important results should be verified by a human. (ai.baidu.com)
Pricing
As of 2026-11-20, Baidu's official domestic and international pricing materials have been searched, but the input and output unit prices for ordinary API calls to this model have not been verified; no RMB or USD amounts are fabricated. The domestic Agent development billing documentation separately lists limited-time free distribution: when this configuration is enabled on designated channels such as Baidu Search, the web version, the app marketplace, or WeChat Mini Programs, the platform covers this model's resource consumption; QPS is not guaranteed for shared resources. This is not a general price for the ordinary API. (cloud.baidu.com)