Baichuan AI

Baichuan4-Turbo is positioned as a general-purpose model optimized for enterprise high-frequency scenarios, with claims of improved usability versus Baichuan4, and lower deployment/inference cost.

Baichuan4-Air is positioned as a lower-cost, faster general-purpose option using a MoE architecture approach, targeted at cost-sensitive enterprise deployments.

Baichuan4 is positioned as a flagship general-purpose model supporting “search agent” behavior, long-window capability, and multimodal use.

Baichuan-M3 is a medical-enhanced large language model presented as open and commercially usable, designed for clinical assistance with capabilities focused on reliable clinical inquiry behavior.

Baichuan-M2 is a medical-focused model presented as open and commercially usable, with associated technical materials describing training and a verifier-oriented approach.

Baichuan2 is a multilingual general model family presented with technical reporting and open releases across base/chat variants.

Baichuan’s platform site describes a model-service and agent/platform approach for enterprise integration, including a pricing page for model token calls and a set of scenario/workflow examples.

“百小医” is presented on Baichuan’s site as an AI patient companion experience aimed at providing ongoing support scenarios; the site includes product terms and agreement pages for the app and service.