Public Cloud Platforms In China

The newest edition of this evaluation has been out since Q3 2024: The Forrester Wave: Public Cloud Platforms In China, Q3 2024. Its verdict is written in the score counts. Eight Chinese providers, thirty two criteria, and one vendor took twenty three highest scores with the report's own conclusion that it "stands out as a public cloud platform driven by AI innovation."

The report that is already out: The Forrester Wave: Public Cloud Platforms In China, Q3 2024

The China edition was led by VP, principal analyst Charlie Dai and published alongside the global and European editions in the same 2024 wave cycle, a three-part scoring of one market from three vantage points.

The China scorecard evaluates eight domestic providers on thirty two criteria spanning strategy, product, and market performance. The field is the Chinese public cloud itself: Alibaba Cloud, Tencent Cloud, Huawei Cloud, Baidu, and the rest of the domestic market, scored by rules that differ from the global edition because the market itself differs, regulation, data residency, and an AI supply chain with its own foundation models.

Alibaba Cloud led the market with twenty three highest scores and the Leader placement. The citation names strong cloud infrastructure, data management, containers, serverless, and popular generative AI services including Tongyi Lingma, its AI coding assistant, and low-code assistants.

The word worth pausing on is driven. "A public cloud platform driven by AI innovation." Not a cloud with AI features, but a cloud whose innovation engine is AI itself, and the scorecard grades the providers on how completely they have rebuilt around that engine.

The field composition is its own finding. Eight domestic providers, no foreign hyperscalers, because the China market's rules, data residency, licensing, model supply chain, make it a parallel market rather than a regional slice of the global one. A buyer who benchmarks Chinese vendors against the global edition's criteria is comparing across two different rulebooks, and the existence of a separate China scorecard is the market's own acknowledgment of that distance.

The serverless note the analysts flagged

The global takeaways blog, written by Sustar, Dai, and Maisto, singles out the China market for one observation the rest of the world should copy: Chinese providers are particularly innovative in serverless machine learning and deep learning, and mini-app mobile support.

The observation is the market's structure in miniature. China's cloud market skipped some of the infrastructure generations the West went through, and its providers built serverless-first in places where the Western giants still run container-first. The scorecard's serverless criteria reward that pattern, and Alibaba's twenty three maximum scores include serverless and FaaS, the same area where it took maximum scores in the global edition.

A buyer evaluating the China market should carry that distinction. The China scorecard is not a smaller version of the global one. It is a different market with a different architecture gradient, and the provider rankings follow the gradient.

The AI layer that is the real differentiator

The China edition's AI criteria are the market's actual battleground, because the foundation model supply chain in China is domestic. Tongyi Lingma, Alibaba's coding assistant, scores alongside low-code assistants and model services, and the report's framing, AI-driven innovation, means the evaluation measures how fast each provider turns its models into usable services.

The global edition's caution applies here with extra weight: many AI services across the whole cycle were not yet generally available at evaluation time. In China's market, where the model landscape moves on a domestic release cadence, a Q3 2024 AI score is a snapshot of a faster-moving object than most.

The same vendor also exposes the editions' disagreement in public. Alibaba Cloud leads the China edition and holds a Leader seat in the global Q4 2024 edition, with seventeen maximum scores there, yet the two scorecards grade different capabilities against different market rules. The dual result is the cleanest available demonstration that cloud leadership is now jurisdiction-specific, and that a vendor's tier in one market is evidence about that market only.

The honest limits of the record

One Leader placement and the field's shape are publicly confirmed; the full tier table has not surfaced in the open record, and this article assigns no placements the record does not support.

There is also the triad to remember. The China scorecard exists inside a three-edition cycle, and its verdicts should be read against the global and Europe editions, not alone. The same vendor can lead in China, take a Leader seat globally, and face a different set of questions in Europe, and the three scorecards are the market saying that cloud leadership is now jurisdiction-specific.

For a buyer, that means the China tier column is evidence about China. It is not a proxy for global capability, not a guide to European compliance, and not a shortcut around the other two editions, and any vendor pitch that treats the three scorecards as interchangeable is pitching harder than it is scoring.

Three questions for the China cloud buyer

Which scorecard applies to your workloads: China, global, or both? If your estate spans jurisdictions, the three editions disagree about the same vendors, and your procurement is three procurements. Match the scorecard to the jurisdiction before you match the tier to the vendor.

Is the vendor's AI actually generally available? The cycle's own caveat says many scored services were not. For every AI service in the pitch, demand general availability dates and reference deployments in your industry, and discount the rest.

Where does your data sit, and which models touch it? The China market's data residency and domestic model supply chain make the architecture question a compliance question. Ask where the training, the inference, and the data all physically happen, and get it in the contract.

Forrester does not endorse any vendor named here, and tier placement should not be read as a recommendation to buy.

Worth reading alongside Private 5G Services: the first ever scorecard for this category named a neutral network host, not an equipment vendor or a telecom giant, as its Leader, evidence the market's value has moved from the radio itself to whoever operates everything riding on it.