Large Model Concept Stocks Volatility: Intensive Product Releases and Comprehensive Multi-functional Application Landscape Unfold
Recently, enterprises related to large language models have entered a period of intensive product releases. A multi-polar, multi-functional application landscape is unfolding across the board.
On May 20, 2026, at the Google I/O developer conference, the company released Gemini 3.5 Flash (positioned for high-concurrency, low-cost real-time interaction, with an inference speed reaching 284.2 tokens/s, approximately four times that of GPT-5.5), the omni-modal model Gemini Omni (possessing physical consistency and capable of generating 10-minute long videos), and the all-weather AI agent Gemini Spark.
On May 21, 2026, Zhipu announced that it had recently partnered with Yuxun Network and Tsinghua University to successfully implement its next-generation network architecture, ZCube, at scale within the GLM-5.1 online production cluster. It also announced the availability of the "GLM-5.1-highspeed" API for selected enterprise clients, with a model output speed reaching 400 tokens/s, setting a new global benchmark for API speeds among large model vendors.
Recently, Mininglamp Technology released Octic, its first AI-native recording hardware. As the first hardware gateway for its AI-native platform, it targets meeting scenarios. Through the edge model Mano-P, the open-source agent collaboration platform Octo, and the Octic hardware carrier, it forms a complete closed loop from perception to processing and collaboration. Furthermore, Mininglamp Technology CEO Wu Minghui revealed that the company has achieved near-simultaneous online collaboration among nearly 1,000 employees and over 1,500 AI Agents. The MOA (Mixture-of-Agents) architecture forms a transparent and traceable "white-box" system through collaboration between agents.
In addition, the recent intensive releases of AI product matrices by major tech giants such as Tencent Cloud, Alibaba Cloud, and Baidu clearly outline a realistic path for AI applications to move from "proof of concept" to "large-scale deployment."
If capital and computing power are the two wheels of the AI industry, then the widespread blossoming of applications is the ultimate engine driving the realization of industry value. The "entry wave" of AI applications has erupted across the board, showing strong performance ranging from Agents to AI Coding, and from major tech giants to vertical scenarios.
Catalyzed by industry and company news, stocks related to the large model concept have performed well this week. As of the market close on May 22, Zhipu saw a weekly gain of over 20%, Mininglamp Technology-W rose over 10% for the week, and stocks such as MINIMAX and Xunce experienced significant rebounds on May 22.
Industry analysis suggests that the tipping point for the AI application era has arrived. Kaiyuan Securities stated that multi-modal large models can deeply empower content production, marketing, and industrial manufacturing. As long as the value created by AI in the workflow exceeds the cost of tokens, there is no ceiling on demand. In 2026, the call volume for tokens in leading models jumped, and the commercialization space for domestic models remains vast.
The capital market is no longer looking only at "whether they can build a model," but has begun to look at "whether it can be transformed into assets." The Chinese large model industry is moving from the excitement of the technology circle into the high-stakes arena of the capital market.
Currently, different large model-related enterprises are engaged in differentiated competition. For instance, Zhipu takes model intelligence as its core driving force and is experiencing an acceleration in commercialization brought about by the explosion in API demand. MiniMax is using a dual-drive strategy of globalization and full-modality to transition toward a platform company. Mininglamp Technology is extending from the B-end to the C-end by deepening its footprint in Agentic AI niche scenarios, driven by the DeepMiner large model product line, models such as Mano/Cito, and the Cider inference acceleration framework. Xunce has carved out a position based on its "vertical token data hub," transitioning from a traditional subscription model to a token-payment and value-sharing model, advancing in scenarios such as finance, robotics, and commercial aerospace. Different enterprises are entering from different dimensions, all pointing to the core proposition of AI competition.
As China Merchants Securities pointed out, model intelligence, multi-modal capabilities, and architectural efficiency constitute the three core dimensions of competition in the large model industry. Kaiyuan Securities stated that commercialization will be a critical theme for large model companies in 2026; those who can convert token consumption into sustainable revenue growth will occupy a more favorable position in this wave of AI.
|
Hong Kong-listed Large Model Concept Stocks: |
|
|
KNOWLEDGE ATLAS (02513.HK) |
Knowledge Atlas Technology Joint Stock Company Limited (02513.HK; commercially known as Zhipu AI) is a leading player in China's independent large language model (LLM) sector (recognized for its capabilities and market position by Frost & Sullivan). The company has released its next-generation flagship model, GLM-5, achieving open-source State-of-the-Art (SOTA) performance in coding and agentic capabilities. It has also open-sourced the multimodal image generation model GLM-Image in collaboration with Huawei. Focusing on novel model architecture design, generalized reinforcement learning paradigms, and autonomous model evolution, its business layout closely aligns with the trend of enterprise-level AI productivity transformation. |
|
Mininglamp Technology-W (02718.HK) |
As a large model concept stock, Mininglamp Technology is recognized by the market as the "first Agentic AI stock on the Hong Kong Stock Exchange." Its self-developed model portfolio includes the DeepMiner large model product line (the underlying engine), models such as Mano/Cito, and the Cider inference acceleration framework. All these are interconnected via the Octo platform layer, which serves as a central hub for human-Agent collaboration. Ultimately, they are delivered in the form of Agentic Services, empowering the implementation of decision-making AI agents in industries such as marketing and mass consumer goods.Rather than pursuing the "Scaling Up" path of monolithic large models, the company adopts a "Scaling Out" approach through the collaboration of multiple specialized small models, achieving accuracy that surpasses general-purpose models in vertical scenarios. Its core competitive barriers lie in niche scenario data, specialized models, and continuous learning. Furthermore, the company has built an open-source, private-deployable, and white-box auditable Private AI infrastructure. |
|
XUNCE (03317.HK) |
A leading data infrastructure provider for financial asset management in China, recognized by the market as the "First Token Stock." The company continues to deepen the integration of LLMs into the core business workflows of financial asset management, reconstructing investment research, trading, and risk control pipelines through underlying data governance and AI technologies. Its underlying structured data cleaning capabilities closely align with the digital transformation of financial institutions and the compliant implementation trends of LLMs. |
|
MINIMAX-W (00100.HK) |
A top-tier player dedicated to omni-modal Artificial General Intelligence (AGI). The company continuously iterates its video generation model matrix and has partnered with mainstream domestic and overseas chip manufacturers and inference platforms to complete underlying adaptations. The total monthly active users (MAU) of its B2C and B2B LLM applications, along with multimodal API call volumes, maintain rapid growth, further broadening its computing power ecological layout. |
NewTimeSpace Disclaimer: All content herein is the original work of NewTimeSpace. Any reproduction, reprinting, or use of this content in any other manner must clearly indicate the source as "NewTimeSpace". NewTimeSpace and its authorized third-party information providers strive to ensure the accuracy and reliability of the data, but do not guarantee the absolute correctness thereof. This content is for reference only and does not constitute any investment advice. All transaction risks shall be borne by the user.
- Large Model Concept Stocks Watch | Zhipu (02513.HK) Surges Over 23% Intraday, Refreshing All-Time High Since Listing
- Huaan Hang Seng Internet Technology Industry ETF(159688) Rises 1.09%; Large Model-Related Enterprises Trend Higher
- Large Model Concept Stocks Watch | Mininglamp Technology-W (02718.HK) Continues to Rise in Morning Trading, Cumulative Weekly Gains Exceed 10%
- ETF Financing Ranking | Guotai CSI Semiconductor Material Equipment Theme ETF (159516) had a net financing buy-in of CNY 121.08 million on may 20, ranked first among stock ETFs
- HUISHENG INTL (01340.HK) plans to issue up to 184.57 million shares by placement, with estimated net proceeds of HKD 8.22 million
