Asian startups raised US$38.5 billion in seed through growth-stage funding in the third quarter, according to Crunchbase data, keeping venture investment in the region near its strongest levels in years even as the money gathered around a small group of artificial intelligence developers, most of them in China.

Crunchbase, which published its tally on Oct. 16 using data collected through Oct. 5, described the total as the second-highest in years, though down from the second quarter. That earlier period was lifted by a $7.4 billion financing for DeepSeek, the Chinese AI model developer. The number of deals was relatively flat from the previous three months.

About 55% of the quarter's funding, an estimated $21 billion, went to AI-focused companies, Crunchbase said. For investors in the region, the figures describe a market where capital is abundant for perceived leaders and harder to find for everyone else.

China Takes Seven of Every 10 Dollars

China accounted for about $27 billion, roughly 70% of the Asian total. India came second with $2.9 billion and Singapore third with $2.5 billion, and Crunchbase said both were down significantly from the second quarter.

The two largest rounds went to AI companies with roots in China, according to Crunchbase. Moonshot AI, a Beijing-based foundation model developer, raised a $3.5 billion Series F in July at a reported valuation of $35 billion. Kling AI, the video-generation business spun out of China-based short-video platform Kuaishou, raised about $3 billion in a July financing that included Baidu and Tencent.

Robotics also drew large checks. Shenzhen-based D-Robotics raised a $400 million Series C in September. In Hong Kong, e-commerce finance company Qupital took in a $300 million Series C as it eyes an initial public offering, Crunchbase said.

Bigger Rounds, Later Stages

Late-stage rounds, which Crunchbase defines as Series C and beyond, totaled about $20 billion. That was slightly below the second quarter but nearly double the level of a year earlier.

Early-stage funding, covering Series A and B rounds, fell 28% from the second quarter to $13.6 billion, though it remained 65% higher than a year before. Seed and angel investment rose about 22% quarter on quarter to $4.9 billion, more than twice the year-earlier amount.

Other data providers show the same pull toward very large deals. Tracxn counted $13.5 billion of tech funding in Southeast Asia in the first nine months of 2026, up 176% from a year earlier, even though the number of rounds fell 28% to 189, TNGlobal reported. Four rounds of $1 billion or more accounted for 63% of the region's total.

Tracxn's figures also show how classification can move regional totals. It counted Kling AI's $2.8 billion July Series D as a Singapore deal and described the company as Singapore-based. Crunchbase described Kling AI only as a spinout of China-based Kuaishou and did not say which country's total included the round.

In India, Tracxn data reported by Analytics India Magazine showed tech funding up 7% to $10.3 billion over the first nine months, while the number of funding rounds dropped 38% to 1,134. Seed funding fell 37% to $698 million over the same period, a contrast with the quarterly rise in seed money that Crunchbase found for Asia as a whole.

What to Watch

The concentration matters beyond private markets. When heavily funded companies move toward listings, their latest private valuations become the reference point for public investors. Qupital is eyeing an IPO, according to Crunchbase.

The quarterly figures are also sensitive to single transactions. The DeepSeek financing alone supplied $7.4 billion of the second quarter's total, and Moonshot AI and Kling AI together accounted for about $6.5 billion of the third. A quarter without deals of that size could show a steep drop in headline funding with little change in the broader pace of investment.

For that reason, deal counts and the mix of stages are as useful as the totals. On those measures, the third quarter showed rising seed funding, a pullback in Series A and B funding and continued heavy spending on late-stage AI companies. Fourth-quarter figures will show whether the run of very large rounds carries into 2027.