Chinese AI company Baichuan says the latest version of its large language model has a larger "context window" than its overseas competitors.
Chinese AI company Baichuan also claims its model outperforms Anthropic's Claude 2 in response quality as well as in understanding long text.
Chinese AI company Baichuan has launched an AI model that it says can digest and summarize novels, making it the world's most powerful model for processing long text prompts.
The Beijing-based company, founded by Chinese search engine Sogou founder Wang Xiaochuan, announced its latest version of the Baichuan2-192k Large Language Model (LLM) on Monday, saying its "context window" can handle about 350,000 Chinese characters.
A context window is a combination of input and output text that an AI model can process during conversations with users.
By comparison, Claude 2, unveiled in July by Amazon.com's Anthropic as angola email address the world's most advanced AI model in terms of the number of words users can include in their chat queries, is said to have a context window of about 75,000 English words, corresponding to hundreds of pages of documents or a book.
According to a report by Chinese company WeChat, the context window of the Baichuan model is 14 times larger than that of OpenAI GPT-4-32k.
Read more CHINA NEWS Chinese economy maintains recovery momentum
Chinese AI company Baichuan also said its model outperformed Claude 2 in answer quality and in understanding and summarizing long text, citing results from LongEval, a project launched by the University of California, Berkeley and other US institutions to assess how well LLMs handle large prompts.