
Xihu Da Model is self-developed by Xihu Xinchen (Hangzhou) Technology Co.Universal Large Model, which is described in detail below:
Background and basic information on the launch
- Launch time: July 9, 2023
- affiliated company: Xihu Xinchen (Hangzhou) Technology Co.
- creator: Zhenzhong Lan (Doctoral Supervisor, Westlake University)
- R&D HistoryAfter graduating from Carnegie Mellon University (CMU) with a Ph.D., Zhenzhong Lan was a scientist at the Google Artificial Intelligence Research Institute, where he successfully developed a lightweight version of Google's big model, BERT, called "ALBERT. 2020, Zhenzhong returned to China and joined Xihu University, where he began to research and develop his own general-purpose big model. In 2020, Lan Zhenzhong returned to China and joined Westlake University to start the self-development of the universal big model. After two years of dedicated research and development, the Xihu Xinchen team has constructed a multimodal "Xihu Big Model".
Model Features
- multimodal capability: The Westlake Grand Model incorporates multiple modalities such as text, image, speech, and vision, and the modalities will be deeply integrated with each other so that the model can collectively obtain information from multiple modalities for more in-depth reasoning.
- Long-term memory and emotion perception: The model is able to remember the user's historical interaction information, deeply customize the model's persona, personality and preferences, and can sense the user's emotional state and adopt a warm expression to interact with the user.
- High IQ and high EQ: The Westlake Big Model has ranked high in authoritative assessments such as SuperCLUE and C-EVAL, demonstrating excellent intellectual and reasoning abilities. At the same time, its emotional perception and long-term memory ability also make it characterized by high emotional intelligence.
Application Scenarios and Results
- Gaming: Westlake Big Model cooperates with Jinke Tomcat to apply AI technology to game characters, providing players with a smarter and more interesting interactive experience. Tomcat has about 400 million monthly active users globally, and it is expected that the number of model calls per year can reach hundreds of billions after full access to Westlake Big Model.
- Clothing design field: Westlake Heartstar and Zhiyi Technology, which is deeply involved in the apparel industry, have jointly launched a garment design industry-orientedAI macromodel--Fashion Diffusion: Designers simply describe the features of a garment and the model generates a high-quality, ready-to-wear image of the garment in less than 10 seconds.
- Psychological counseling fieldWestlake Heart launched the AI psychological counseling app "Chatting with Xiaotian", which provides instant, effective, low-cost and warm AI psychological counseling services to the public through virtual counselors.
Market position and influence
- user basis: Less than a year after the launch of the Westlake Big Model, it has accumulated more than two million C-end users and served hundreds of B-end enterprises, including Tomcat, Know Your Clothes, Starbucks, Alipay, Know Your Clothes, Cool Joy and other well-known enterprises.
- Financing: Xihu Xinchen has received a new round of over 100 million RMB strategic investment from Tomcat, and the two sides will accelerate the landing of the application of big models. In addition, Xihu Xinchen has also received ten million dollars of investment from well-known organizations such as Bluechip Venture Capital, Baidu Ventures, and Kaitai Capital.
- Policy filing: The Westlake Grand Model has been filed through the Interim Measures for the Management of Generative Artificial Intelligence Services in December 2023, opening up related services to the public for the whole society.
Xihu Xinchen will focus on the research and development of multi-modal large models and the creation of industrialized applications, and is committed to constructing a multi-modal intelligent body platform that can be deeply customized, which widely serves the fields of finance, e-commerce, gaming, social networking, education, medical care, and recreation. In the future, the application scenarios of Xihu Big Model will be further expanded to provide intelligent solutions for more industries and fields.
In summary, as a self-developed general-purpose big model with outstanding performance in multimodal ability, long-term memory and emotion perception, high IQ and high EQ, the Westlake Big Model has achieved remarkable results in a number of fields and has broad market prospects and influence.
data statistics
Relevant Navigation

The next-generation multimodal big model introduced by OpenAI has stronger language understanding, generation and cross-modal interaction capabilities, and is widely applicable to content creation, intelligent assistants and enterprise applications.

ERNIE
Baidu's industrial-grade knowledge-enhancing big models, with industry-leading natural language understanding and generation capabilities, are widely used in all kinds of natural language processing and generation tasks, helping enterprises realize intelligent upgrading.

DeepSeek-V3
Hangzhou Depth Seeker has launched an efficient open source language model with 67.1 billion parameters, using a hybrid expert architecture that excels at handling math, coding and multilingual tasks.

Wan2.1
Alibaba launched an efficient video generation model that can accurately simulate complex scenes and actions, support Chinese and English special effects, and lead a new era of AI video creation.

Zidong Taichu
The cross-modal general artificial intelligence platform developed by the Institute of Automation of the Chinese Academy of Sciences has the world's first graphic, text and audio three-modal pre-training model with cross-modal comprehension and generation capabilities, supporting full-scene AI applications, which is a major breakthrough towards general artificial intelligence.

WebLI-100B
Google DeepMind launches a 100 billion visual language dataset designed to enhance the cultural diversity and multilingualism of AI models.

HunyuanImage2.1
Tencent launched the open source raw image model, which natively supports 2K HD raw images, accurately parses complex semantics, and can efficiently generate high-quality images with Chinese and English fusion.

Claude 3.7 Sonnet
Anthropic has released the world's first hybrid reasoning model that demonstrates superior performance and flexibility by being able to flexibly switch between rapid response and deeper reflection based on different needs.
No comments...
