Data to be translated: BaiChuan Intelligence has released the Baichuan2—Turbo series of APIs, supporting a super-long context window of 192K and enhancing the capability of search-enhanced knowledge bases. Users can create their own knowledge bases by uploading text materials, achieving more comprehensive and efficient intelligent solutions. The official website model experience has also been upgraded, supporting multiple text uploads and URL inputs. BaiChuan Intelligence believes that search enhancement is the key to large model applications, capable of addressing issues such as hallucinations, poor timeliness, and insufficient knowledge in specialized fields. By combining large models with search enhancement technology stacks, links between large models and domain knowledge as well as internet-wide knowledge are realized. Enterprises can utilize large model + search enhancement solutions to customize their knowledge base needs, improving the usability and stability of the models. BaiChuan Intelligence has overcome multiple challenges in search enhancement technology, enhancing the recall rate and accuracy of the models. Through parallel sparse retrieval and vector retrieval, a 95% accuracy rate in responses can be achieved. BaiChuan Intelligence's search-enhanced knowledge base also leads in industry-leading model scores across various test sets.