Huawei Cloud has officially accelerated the regional AI race by launching its Model-as-a-Service (MaaS) in Asia Pacific. Announced during the “Agentic AI Practice” summit in Jakarta on 15 April 2026, the service introduces a high-quality Token Service powered by Huawei’s in-house acceleration engine to support mainstream open-source models.
The launch signifies a transition from providing raw AI compute to offering a full-lifecycle model management platform, allowing enterprises to deploy, fine-tune, and evaluate AI with “one-click” ease and pay-per-use billing.
GLM-5 and the Rise of Enterprise Agents
A major highlight of the rollout is the support for GLM-5, a model that has achieved state-of-the-art (SOTA) performance in coding and agentic capabilities.
- Target Scenarios: The MaaS platform initially focuses on two primary applications: Intelligent Q&A and AI Coding.
- Model Ecosystem: Beyond GLM, Huawei is supporting a diverse library including DeepSeek and Qwen categories, ensuring developers have access to the industry’s highest-quality open-source foundations.
- The “Arts” Roadmap: While MaaS and ModelArts are live, Huawei confirmed that CodeArts (coding agent) and AgentArts (agent platform) will launch outside the Chinese mainland in the second half of 2026.
Infrastructure & Adoption in ASEAN
Huawei has optimized its regional cloud network to ensure a 50 ms access latency across key markets, including a unique “local cloud” partnership in Malaysia.
| Partner Case Study | Industry | AI Implementation Outcome |
| iFLYTEK (Singapore) | Intelligent Voice / AI | Deployed LLM training resources in two weeks; achieved 60 days of uninterrupted training stability. |
| GSM (Vietnam) | Green Mobility | Utilised IoTDA and ModelArts to analyse violations and generate real-time driver/passenger risk alerts. |
| Regional Network | Cloud Infrastructure | 5 Regions and 18 Availability Zones (AZs) spanning Singapore, Thailand, HK, Indonesia, and the Philippines. |
Editor’s Take: From “Migrating to Cloud” to “Using AI Well”
For the Malaysian Business reader, Huawei’s MaaS launch is a pivotal moment in the commoditization of AI. By shifting to a Token Service model, Huawei is lowering the barrier to entry for Malaysian SMEs and enterprises who may not have the capital for heavy GPU investment but need the competitive edge of Agentic AI. As GSM’s success in Vietnam shows, the real value lies in integrating AI into the core “Value Chain” (the ACES 4Cs) to drive safety and operational efficiency.