Infinigence AI has completed a RMB 2.3 billion Series C round — about $344 million — taking its total funding to RMB 4.3 billion since 2023. The Shanghai company builds the software layer that lets AI models run efficiently across many different chips at once, and that flexibility is becoming strategically valuable in China, where access to the most advanced US processors is restricted.
October 12, 2026 — Infinigence AI announced the completed Series C through company materials released on Monday, in what Chinese reporting describes as a record raise for the country’s AI-native infrastructure category.
What happened
Infinigence AI, founded in 2023, has closed a RMB 2.3 billion (roughly $344 million) Series C funding round. The deal brings the company’s cumulative financing to RMB 4.3 billion, or about $644 million, in just three years. Reporting on the round names a broad investor roster spanning state-backed funds and industrial backers, including Tencent, Baidu, Lenovo, Sequoia China, Qiming Venture Partners and Shunwei Capital, although the exact allocations for this tranche were not separately disclosed.
What Infinigence actually sells
The company develops infrastructure software that allows AI models to run efficiently across different chip platforms instead of being tied to a single hardware family. Its platform connects 37,000 petaflops of computing capacity across 16 mainstream chip types and supports mixed-chip training using both Chinese and international processors. On top of that base, Infinigence sells model inference services and infrastructure for enterprise AI agents, with deployments in gaming, healthcare, legal services and energy.
The growth numbers are company-reported, so treat them as vendor figures: Infinigence says daily token usage on its model-service platform grew 40-fold by July compared with December 2025, and that its cross-cluster inference architecture cut time-to-first-token by 51.5% and cost per token by 37.5%.
Why it matters
Infinigence is a clean example of where AI money is going in China. US export controls mean the most advanced Nvidia chips cannot be sold there, so squeezing serious work out of mixed fleets — domestic processors like Huawei’s Ascend line alongside international hardware — is a necessity, not a niche. A company that makes heterogeneous compute behave like one platform sells exactly the workaround the market needs, and it counts major Chinese model developers among the users of that approach.
The funding also looks like a runway to a listing. Bloomberg reporting carried by Tech in Asia says Infinigence confidentially filed for a Hong Kong IPO in early October, aiming to raise several hundred million dollars with a debut possible as early as the first half of 2027, at a reported pre-IPO valuation of RMB 14.3 billion. None of that timetable is guaranteed, but a $344 million Series C followed by a public listing would make Infinigence one of the defining tests of investor appetite for China’s AI infrastructure layer.
Frequently asked questions
How much has Infinigence AI raised in total?
RMB 4.3 billion (about $644 million) since its 2023 founding, including the RMB 2.3 billion Series C completed in October 2026.
What does Infinigence AI do?
It builds AI infrastructure software that runs models across 16 chip types — 37,000 petaflops of connected capacity — plus inference services and tooling for enterprise AI agents.
Is Infinigence AI going public?
Reporting says it confidentially filed for a Hong Kong IPO in early October 2026, with a listing possible in the first half of 2027. The company has not confirmed a date.
Sources: TechNode, Tech in Asia, PANews, TechStartups