Flowith Raises Tens of Millions in Seed Round
By Chen Wei
Image / Photo by Ant Rozetsky on Unsplash
Flowith just raised tens of millions to turn AI into action on the factory floor.
Flowith disclosed a multi-million dollar seed and seed+ round that, together, totals in the tens of millions of U.S. dollars. The seed was backed by Vertex Ventures, while the seed+ round was jointly led by Sequoia China Seed Fund and LongRiver Investments, with other institutional investors contributing as well. The funding surge aims to accelerate R&D and global market expansion for Flowith’s evolving “Agentic AI” stack. In.flowing shorthand, Flowith wants to move from chat to concrete action, and the math behind that is as much about systems as it is about smart software.
At the core, Flowith frames the shift as an evolution from conversational AI to action-ready AI—what it calls an OS-level agent shift. The company has already rolled out a set of products that map to this ambition: an open, general-purpose creative agent framework called Oracle; an AI Context “Knowledge Garden” to organize data context; an unlimited-step agent named Agent Neo; and FlowithOS, an OS-level agent system designed to orchestrate tool use and workflow execution. Flowith says these tools have attracted millions of active users globally, a signal that the market is hungry for end-to-end automation that can bridge planning with execution.
The timing matters. Flowith’s emphasis on system-level agents comes as Openclaw and other demonstrations in 2026 highlighted the potential for AI agents to perform autonomous tasks, not just draft text. In practical terms, Flowith is betting that enterprises will want AI to plan, decide, and act—without human-in-the-loop prompts at every step. That demands robust tool invocation, safe fallbacks, and a governance layer to prevent unwanted outcomes. In China and other large manufacturing ecosystems, that is especially salient: OS-level agents must co-exist with industrial control systems, ERP/SCADA integrations, and strict data-security regimes.
From a policy and market perspective, the funding signals a continuing push from top-tier capital into AI platforms that aim to be underlying infrastructure rather than single-use apps. The investor lineup—Vertex Ventures, Sequoia China Seed Fund, LongRiver Investments, and others—reads as a blend of global strategic capital and deep Chinese market access. Flowith’s explicit aim to fuel “global market expansion” alongside R&D suggests a two-track path: build a scalable, cross-border software stack while tailoring components to domestic manufacturing environments where data localization and cybersecurity standards are non-negotiable.
Two practitioner takeaways stand out for supply chain and manufacturing executives watching this space. First, expect a tension between flexibility and control. Flowith’s OS-level agents promise powerful automation, but manufacturing buyers will demand clear governance, safety rails, and predictable performance when an autonomous agent starts controlling toolchains or production workflows. Second, data strategy will be decisive. Factory data is fragmented across MES, ERP, PLCs, and robotics, with sensitive informational needs and strict retention rules. Any adoption of FlowithOS or its agent frameworks will require careful data localization, secure integration into existing networks, and explicit fallback modes if external agents lose connectivity.
For Chinese manufacturers and global players competing with China, Flowith’s seed round underscores a broader trend: the rise of system-level AI platforms as infrastructure for automation. If Flowith can demonstrate real-world productivity gains—shorter changeovers, smarter maintenance planning, and tighter coordination across supply chains—the model could accelerate the adoption curve for AI-enabled manufacturing tools across provinces and industry clusters. It also means a continued duet of private capital and domestic market demand shaping the next wave of industrial AI, even as global regulators press for safety and data governance.
Key Chinese terms translated with context: 代理AI (agentic AI) refers to AI that can autonomously act, not just generate text; 操作系统级别的智能代理系统 FlowithOS translates to an OS-level intelligent agent system designed to orchestrate tool use and workflows; 知识花园 (Knowledge Garden) is Flowith’s data context organization concept; 甲骨文/Oracle here denotes Flowith’s general-purpose creative agent framework, a term borrowed from the company’s product naming.
Ownership remains described as private venture-backed, with no public indication of state-owned backing. The presence of Sequoia China Seed Fund as a lead investor alongside Vertex Ventures and LongRiver Investments signals strong private-sector confidence in Flowith’s platform approach—and a potential signal to factory-floor software buyers that the AI operating system layer is moving from theory to deployment.
Sources
Newsletter
The Robotics Briefing
Weekly intelligence on automation, regulation, and investment trends - crafted for operators, researchers, and policy leaders.
No spam. Unsubscribe anytime. Read our privacy policy for details.