Flowgram.Ai
Flowgram.Ai is an extensible workflow development framework with built-in canvas, form, variable, and materials that helps developers build AI workflow platforms faster and simpler.
Last verified:
What is Flowgram.Ai?
Flowgram.Ai is an open-source, extensible workflow development framework and toolkit designed to help developers build AI workflow platforms faster and simpler. It provides a composable, visual, easy-to-integrate solution with built-in tools including a flow canvas, node configuration form engine, variable scope chain management, and ready-to-use materials like LLM nodes, Condition nodes, and Code Editor nodes. Unlike ready-made workflow platforms, Flowgram.Ai provides the building blocks to create your own custom AI workflow platform or automation system.
Key features include two canvas modes: Free Layout Canvas where nodes can be placed anywhere with free-form connection lines, and Fixed Layout Canvas where nodes are dragged to specified positions with support for compound nodes like branches and loops. The form engine manages CRUD operations for node data with rendering, validation, side effects, linkage, and error-capturing capabilities. The variable engine supports scope constraints, variable structure inspection, and type inference for managing data flow within workflows. The toolkit also provides pre-built materials (components, side effects, validators) that developers can quickly reuse and extend.
Flowgram.Ai is specifically designed for developers building AI workflow platforms, automation systems, and no-code/low-code platforms. It was created based on ByteDance's internal needs for building diverse AI workflow platforms with complex business logic. The tool is used by products like Coze Studio (AI agent development tool), NNDeploy (multi-platform AI deployment tool), and Certimate (open-source SSL certificate management tool). It's particularly suitable for visual workflows with clear inputs and outputs, and supports both TypeScript and JavaScript development.
Flowgram.Ai pricing
Pricing model: Freemium
FlowGram is open-source software released under the MIT license, which means it is completely free to use. There are no paid tiers or subscription plans mentioned on the website. Developers can create new FlowGram projects using npx @flowgram.ai/create-app@latest at no cost. The framework provides all core features including canvas, form engine, variable engine, and materials without any licensing fees.
Flowgram.Ai pros
- Open-source with MIT license
- 7.1k GitHub stars showing strong community adoption
- Built-in visual flow canvas with two layout modes
- Free Layout mode for flexible node positioning
- Fixed Layout mode with support for branches and loops
- Form engine with validation and error handling built-in
- Variable engine with scope constraints and type inference
- Ready-to-use materials including LLM, Condition, Code Editor nodes
- Easy npm installation with create-app scaffolding
- Supports CodeSandbox and StackBlitz for quick online demos
- Comprehensive documentation with API reference
- Best practice interaction patterns (undo/redo, copy/paste, gestures)
- Motion transition animations for smooth UX
- Mac trackpad gesture support for canvas zooming
- TypeScript-first with 79.3% TypeScript codebase
- Extensible architecture for custom node development
- Used by production tools like Coze Studio and Certimate
Flowgram.Ai cons
- Not a ready-made product - requires development effort to build your platform
- Primarily targets developers, not non-technical users
- Limited pricing information available on website
- Documentation mainly in Chinese with some English sections
- Requires Node.js installation for local development
- Learning curve for understanding canvas, form, variable, and material concepts
- No hosted cloud version - self-hosted only
- Limited pre-built materials compared to commercial workflow platforms
Frequently asked questions about Flowgram.Ai
What is FlowGram?
FlowGram is a composable, visual, easy-to-integrate, and extensible workflow development framework and toolkit. It helps developers build AI workflow platforms faster and simpler by providing built-in tools like flow canvas, node configuration form, variable scope chain, and ready-to-use materials (LLM, Condition, Code Editor, etc.). It is not a ready-made workflow platform but the framework and toolkit to build your own.
Who is FlowGram for?
FlowGram is designed for developers who want to build AI workflow platforms, automation systems, or low-code/no-code platforms. It was created based on ByteDance's internal needs for building diverse AI workflow platforms with complex business logic. It is particularly suitable for developers building visual workflows with clear inputs and outputs.
What are the two canvas layout modes?
FlowGram provides two canvas modes: Free Layout Canvas where nodes can be placed anywhere and connected using free-form lines, and Fixed Layout Canvas where nodes are dragged to specified positions with support for compound nodes like branches and loops. The Free Layout Demo is recommended for quick start.
How do I get started with FlowGram?
You can quickly start by running npx @flowgram.ai/create-app@latest and choosing the Free Layout Demo template. Then navigate to the project folder, run npm install, and npm start to open http://localhost:3000 in your browser. Alternatively, you can try online demos directly in CodeSandbox or StackBlitz without local installation.
What materials (pre-built nodes) are included?
FlowGram comes with ready-to-use materials including LLM nodes for large language model calls, Condition nodes for conditional logic, Code Editor nodes for code execution, and other pre-built node types. These materials can be composed into larger workflows and developers can also create custom nodes and extend the material library.
Is FlowGram open source?
Yes, FlowGram is open-source software released under the MIT license. The GitHub repository is public at github.com/bytedance/flowgram.ai with 7.1k stars and 598 forks. The codebase is primarily TypeScript (79.3%) and the project has 86 releases with active community contribution from 26 contributors.
What programming languages does FlowGram support?
FlowGram is built primarily with TypeScript (79.3% of codebase), with additional JavaScript (2.3%), MDX (17.2%), CSS (0.8%), Less (0.4%), and HTML. It is a TypeScript library that can be used in JavaScript projects as well. The npm packages are available for installation in Node.js projects.
What is the form engine in FlowGram?
The form engine manages the CRUD operations of node data and provides rendering, validation, side effects, linkage, and error-capturing capabilities. This simplifies the development of node configuration forms by handling data management, dynamic form rendering, validation rules, and error handling automatically.
What is the variable engine in FlowGram?
The variable engine supports scope constraints, variable structure inspection, and type inference, making it easy to manage data flow within workflows. It handles variable scope chains, allows developers to inspect variable structures, and performs type inference to ensure type safety in workflow data connections.
Which companies or products use FlowGram?
FlowGram is used by several production tools including Coze Studio (an all-in-one AI agent development tool from ByteDance), NNDeploy (a workflow-based multi-platform AI deployment tool), and Certimate (an open-source SSL certificate management tool for automatic SSL certificate application and deployment with visual workflow).