Chinese open-weight AI models gain global users but lag on revenue

By Splitfeed EditorialLast updated Oct 9, 2026, 10:00 UTC

As of early October 2026, reporting from South China Morning Post and Fortune describes a split picture for Chinese AI startups: their open-weight models are drawing a growing base of global users, but the funding and capital formation those firms command remains well below what their US counterparts have secured. Fortune's framing is direct on the gap, noting that Chinese firms can match US model capabilities technically while falling short on the financial side. A separate analysis published October 9 by the China-Global South Project adds another dimension, arguing that what shapes AI influence in the Global South is not which country produces the most capable frontier model but how deeply a given AI system becomes embedded in local infrastructure and workflows.

The core tension the reporting surfaces is a classic open-source dilemma at geopolitical scale. When a model is released openly, adoption can spread quickly and widely, but the developer does not automatically capture the economic return that adoption generates. Chinese firms appear to be experiencing exactly this: their models circulate, get integrated and get built upon, yet the revenue and valuation gains concentrate elsewhere, plausibly with cloud providers, application-layer companies or developers in other markets who wrap the open-weight models in commercial products. Whether this is a strategic miscalculation or a deliberate play to build influence now and monetise later is genuinely in dispute. The embeddedness argument from the China-Global South Project complicates the picture further, suggesting that if influence rather than revenue is the real prize in developing-market competition, open distribution could still be a winning move even when it does not generate direct income.

None of the three camps, Pro-AI, Anti-AI or Middle Ground, have published tracked reactions to this story yet. Pro-AI voices would typically celebrate the democratising effect of capable open-weight models reaching global users regardless of which company profits, framing wide access as the point. Anti-AI voices would likely focus on the geopolitical risk of widely distributed powerful models with limited oversight, and might question whether the revenue gap signals broader instability in the AI startup ecosystem. A middle-ground perspective would probably centre on the structural question the reporting itself raises, asking whether open-weight release strategies need new monetisation mechanisms before they can sustain the research investment required to keep those models competitive.

The argument will sharpen around a few observable markers: whether any major Chinese AI firm announces a funding round that closes the capital gap Fortune describes, whether cloud or enterprise revenue figures from open-weight model derivatives become public, and whether the embeddedness metrics the China-Global South Project points to, such as integration into government or telecoms infrastructure in the Global South, show measurable movement in the months ahead.