Bubble, a visual development platform that lets users build full web and mobile applications without writing code, just unveiled its Bubble AI Agent in global beta. The new agent allows users to switch between conversational prompts and the platform’s visual editor. According to the company, every agent change is visible, letting builders retain oversight. “Every change Bubble AI Agent makes is visible in our editor,” said co-CEO Josh Haas in the announcement.
The launch is Bubble’s most consequential release since its last major funding round in 2021 and marks its entry into a developer-tool market now defined by AI-driven automation. Bubble calls the new capability “AI speed without compromising control.”
Years of Quiet Foundation
Bubble was founded in 2012 by Emmanuel Straschnov and Josh Haas with the goal of making software creation accessible through a drag-and-drop, logic-based interface. Over time, it evolved into a full-stack platform handling front-end design, workflow logic, database management, and hosting in a single environment.
The company operated bootstrapped for years before raising outside capital. In June 2019 it secured a $6.25 million seed round led by SignalFire, and in July 2021 it closed a $100 million Series A led by Insight Partners with participation from Betaworks and Neo.
The measured pace is unusual in a sector now dominated by aggressive venture funding. In September 2025, Replit raised $250 million at a $3 billion valuation after reporting a revenue surge from $2.8 million to $150 million within a year. In June 2025, Cursor (Anysphere) closed a $900 million Series C valuing it at about $9.9 billion. A month later, Lovable raised $200 million at a $1.8 billion valuation less than a year after launch.
Compared with those figures, Bubble’s $106 million in total funding shows a deliberately conservative approach. Straschnov has said the company focused on building a sustainable platform before pursuing rapid growth. “We bootstrapped for years because we wanted to prove the model worked,” he told Antler Early Days.
By 2025, that foundation supported a large user community. Bubble reported that over 6 million people have built 7 million applications on its platform, processing more than $1 billion in transactions in the past year. Those numbers illustrate a mature ecosystem rather than a beta-stage experiment, which contrasts with many new AI-native builders still searching for scale.
The AI shift forced Bubble to redefine itself. In June 2025 the company introduced its “vibe-code killer” concept and launched a unified web-plus-mobile platform. “The amazing thing about AI is you can express yourself vaguely,” Haas told The New Stack. “But what you’re going to get is millions of lines of AI-written code that isn’t production-ready, isn’t hardened, has bugs and is going to be an unmaintainable nightmare.” That statement summed up Bubble’s philosophy that AI should accelerate development without sacrificing maintainability.
Bringing Discipline to the AI-Coding Rush
The October 2025 AI Agent launch puts that philosophy into practice. The agent listens to natural-language commands and implements changes directly in the Bubble editor. Haas said in the announcement that the goal is to combine “AI speed and complete control over your vision, from design to backend functionality.”
Alongside the agent, Bubble outlined a roadmap for a Security Dashboard (formerly Flusk), a modernized CI/CD pipeline, and built-in in-app purchases for mobile apps. It also announced a $25,000 AI hackathon in partnership with Anthropic to showcase applications built with its agent. The partnership aligns Bubble with a frontier-model developer that also works with Salesforce and Microsoft.
This move places Bubble in a competitive field where both startups and enterprise vendors are redesigning their development tools around AI agents. Microsoft’s Power Platform has added Copilot Studio and agent workflows for its low-code environment. OutSystems, a two-decade-old low-code vendor, launched “Mentor,” a software life-cycle assistant that auto-generates UI and logic from requirements. Another competitor, Creatio, integrated conversational interfaces into its visual tools.
Market watchers see generative AI as accelerating rather than replacing low-code platforms. In a Forrester blog, Principal Analyst John Bratincevic wrote that AI “will quickly evolve the low-code toolset and make the market grow faster.”
Microsoft describes Power Platform as offering “the next layer of abstraction” for professional developers, combining UX, data, and automation in one stack.
Bubble’s hybrid approach fits into that shift. Its visual stack and platform infrastructure serve as a governance layer atop AI-generated changes. Rather than discard visual tools, Bubble overlays agents that produce transparent, editable changes. That sets it apart from pure prompt-to-app tools whose outputs tend to be opaque and harder to audit.
Bubble is not alone in seeing a long-term opportunity here. Enterprise users in regulated industries are demanding accountable AI development systems that log changes and expose logic. By integrating its agent into a visual interface and adding security dashboards and deployment controls, Bubble is aligning with that demand from its existing user base and from new enterprise customers.
Bubble’s AI Agent launch marks a turn in its trajectory. After more than a decade of steady work and measured funding, Bubble is now joining a competitive race on its own terms.
