FlutterFlow Launches AI-Native Builder as Firestore Scale Passes 750 Billion Reads

FlutterFlow has launched an AI-native app builder that turns conversational prompts into cross-platform native iOS and Android applications. As reported by GlobeNewswire, the new product handles application design, backend infrastructure, analytics, and mobile store publishing within chat threads, while the company's existing visual IDE continues under the name FlutterFlow Studio.

The release bifurcates FlutterFlow's product into two distinct development paths: prompt-driven, agent-led generation and granular visual canvas engineering. Co-founder and CEO Abel Mengistu explained that user demand drove the shift to dual modes: "They wanted to prompt for their first idea, and for the changes after. Studio is incredibly powerful, but it didn't give them that experience. We knew that we could build something impactful for this new group of users from the ground up. We also have a fantastic community of millions of builders, built over five years, who want to keep building in Studio. It became clear we needed to build for both."

Turning Prompts into Production Code

Founded in 2020 by former Google engineers Abel Mengistu and Alex Greaves, FlutterFlow originated after the pair encountered the friction of building production client applications during an earlier startup attempt. The platform initially built its base by translating visual layout trees and logic workflows directly into clean Dart source code built on Google's Flutter framework, closing a $25.5 million Series A round reported by TechCrunch from GV, Gradient Ventures, and Y Combinator.

The newer AI-native builder operates through conversational threads where an agent drafts plans, generates widget structures, integrates local state, and inspects visual results. According to a product overview on the FlutterFlow YouTube channel, each task runs on its own isolated copy of the app with dedicated mock data, allowing builders to run multiple agent threads simultaneously without collision before pushing changes into the primary application codebase.

Alongside conversational creation, FlutterFlow has integrated Model Context Protocol (MCP) support into its core editor. Demonstrated in an in-depth walkthrough on the FlutterFlow YouTube channel, developers can connect external coding agents such as Claude Code, Cursor, or Gemini CLI directly to the project canvas, configure project-specific AI skills, trigger automated error fixes from build logs, and roll back project state via discrete checkpoints.

Infrastructure Behind 3.4 Million Builders

While visual and prompt-driven abstractions simplify the developer surface, FlutterFlow's internal architecture must maintain continuous real-time state synchronization across thousands of collaborative editing sessions. According to a case study published by Google Cloud, FlutterFlow has scaled to more than 3.4 million builders across 200 countries, with apps created on the platform serving more than 250 million end users.

The platform relies on Google Cloud's Firestore to coordinate its real-time multi-tenant development environment, processing more than 750 billion reads and 75 billion writes over five years without an outage. In FlutterFlow Studio, real-time listeners track simultaneous canvas changes so designers, engineers, and AI agents can manipulate layout trees concurrently without overwriting code. The case study notes that Firestore documents also serve as the coordination plane for FlutterFlow's Test Mode, streaming hot reloads, build statuses, and debug logs directly between generated Flutter web instances and the browser editor.

Earlier this year at FFDC 2026, the company introduced FlutterFlow 7.0, introducing automated test creation with Test Pilot, native execution of Supabase Edge Functions, and deeper audio model support with speech-to-text and translation.

Engineering and Future Openings

Headquartered in Mountain View with remote hubs in New York and Singapore, FlutterFlow maintains a remote-first team of 55 employees. The company is currently accepting applications across future engineering and general roles.

The team is hiring for Software Engineer - Future Openings on a remote basis globally. Engineers in this role will help build the core application development platform by prioritizing and implementing new features, building custom widgets, developing code generation pipelines, and keeping the platform synchronized with Flutter framework updates. The company is also accepting resumes through its General Application - Future Openings pipeline for upcoming roles across product, engineering, and operations.

See all open roles at FlutterFlow