Dify Secures $30 Million to Help Businesses Deploy AI Agents

The company’s platform helps teams build, deploy and operate artificial intelligence applications and agentic workflows, Dify said in a Monday (March 9) press release. It includes a visual workflow builder with infrastructure for prompt management, tool integration, knowledge retrieval and observability.

Dify will use the new funding to integrate the latest agent capabilities into the core product experience, build a dedicated enterprise product team focused on performance and compliance, launch new initiatives to support builders and deliver product updates, per the release.

“We built Dify so teams can turn their domain knowledge and workflows into production-grade AI applications without rebuilding the stack each time,” Dify founder and CEO Luyu Zhang said in the release. “This funding accelerates our work to make agentic workflows more reliable, more observable and easier for organizations to adopt at scale.”

The funding round values Dify at $180 million, according to the release.

Dify launched its platform in 2023. Today, it runs on over 1.4 million machines worldwide, and commercial versions of Dify are being used by more than 2,000 teams and 280 enterprises, per the release.

Kui Zhou, partner at HSG, which led the round, said in the release: “We believe the AI application layer is at an early but accelerating phase, and platforms that standardize AI development workflows are likely to capture durable value.”

PYMNTS reported in December that enterprise AI is entering a new phase as companies that spent the last two years experimenting with large language models are now moving those systems into live environments.

The PYMNTS Intelligence report “Agentic AI Breaks Out of the Sandbox” found that in the three months between August and November, the share of enterprise chief product officers who refused to grant AI agents meaningful autonomy dropped from 98.3% to 63.3%.

“As 2026 unfolds, the constraint on growth is no longer a matter of technology or interest but of organizational readiness,” the report said. “The ‘just looking’ phase of the AI cycle is winding down, and the winners of the next adoption phase will not be the companies that buy the best models but those that trust them enough to take their hands off the wheel.”

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.