Zhipu.AI Accelerates GLM and Releases as Open Source

Chinese AI company Zhipu.AI has released its large language model 'GLM' series as open source after accelerating its processing speed. The improvements have resulted in a processing speed that is reportedly up to 8 times faster than before. Concurrently, the company launched a new service 'Z.ai' to spearhead global expansion, clearly signaling a full-scale entry into overseas markets.

Zhipu.AI was founded in 2019 based on a research group from Tsinghua University and is recognized as one of the leading developers of large language models within China. The company's GLM series continues to be developed as a model supporting both Chinese and English, and has a track record of contributing to the open source community. This release is positioned as a move to further accelerate these accumulated efforts.

On the technical front, the primary feature is improving the processing speed of the GLM model by up to 8 times. Speed improvements translate directly into benefits for applications and developers actually using the model. As response times shorten, practical utility expands across a wide range of applications, including chatting and real-time information processing.

The newly established 'Z.ai' is expected to function as a platform for overseas users and developers, serving as the gateway for global expansion. While it is not uncommon for Chinese AI companies to launch independent global brands, in the context of current US-China technological tensions, this can be noted as an attempt to gauge how Chinese AI models will be received in international markets.

This series of moves is also discussed in the context of the company potentially considering an IPO (Initial Public Offering). Penetrating the developer community through open source release and enhancing overseas brand recognition through Z.ai can be viewed as a strategy to increase corporate value. However, specific timelines or plans for an IPO have not been confirmed at this point.

From an open source strategy perspective, following Meta's release of the Llama series, a global trend has emerged of providing models freely to attract developer communities. Zhipu.AI's provision of an accelerated version of GLM as open source can be understood as riding this wave while strengthening the international presence of Chinese-developed models.

Key areas to watch going forward are how many overseas users and corporate partners Z.ai can actually acquire, and whether GLM's acceleration will see broad adoption by developers. Zhipu.AI's developments serve as a case study demonstrating how China's AI industry is moving beyond domestic boundaries to participate in global competition, making it worthy of continued attention from an industry-wide perspective.