Giselle
Giselle: AI App Builder. Open Source.
Last verified:
What is Giselle?
Giselle is an open-source AI app builder that enables anyone to intuitively build AI applications through a visual, no-code interface. With drag-and-drop operations, users can create and run AI apps without any programming knowledge. The platform is designed as an execution platform to reliably complete complex, long-running AI tasks as business operations, serving engineers, designers, product managers, marketers, sales, and back-office professionals.
Key features include a Visual Node Editor for designing AI workflows by placing and connecting nodes like a flowchart, Long-Running Task Execution with real-time monitoring in a dedicated environment, and Multi-Model AI Collaboration that freely combines top AI models like OpenAI GPT-5.2, Anthropic Claude Opus 4.5, and Google Gemini 3 Flash within single workflows. GitHub Integration automatically triggers AI apps based on Issues, Pull Requests, and comments for context-aware code reviews and documentation generation. Custom Knowledge lets users incorporate PDFs, documents, and GitHub repository code as AI app knowledge using RAG and Vector Store technology without specialized expertise.
Giselle is purpose-built for AI-native product teams and innovative startups aiming to scale with lean, AI-native teams. It automates repetitive tasks like market research, code review, PRD generation, and documentation updates, freeing teams to focus on high-impact thinking. The platform connects to any foundation model and adapts with the frontier without requiring rewrites.
The tool offers both managed cloud deployment for immediate use and self-hosted deployment options through the open-source version on GitHub. Pre-built workflows include Deep Researcher for AI-powered analysis, PRD Generator for structured product requirements, automated Code Reviewer, and Doc Updater for keeping knowledge bases in sync.
Giselle pricing
Pricing model: Freemium
Free Plan: 30 minutes of AI model usage per month ($3 monthly credit), access to basic AI models, no-code builder, single-user account, managed cloud deployment, community support. Self-hosted open-source version is free with unlimited features including create/deploy agents, multi-agent workflows, and unlimited Document Vector Stores. Pro Plan: $20/month with unlimited AI model usage (limited-time promotional offer), access to all premium AI models, all Free features, $20 monthly AI credits, 5 Document Vector Stores, 3 GitHub Vector Stores, email support, pay-as-you-go additional usage at 10% markup. Team Plan: Coming soon, contact sales for pricing, includes team collaboration up to 10 users, priority email support, 20 Document Vector Stores, 10 GitHub Vector Stores.
Giselle pros
- No-code visual node editor with drag-and-drop interface
- Open-source with Apache License 2.0
- Self-hosted deployment option with unlimited features
- Multi-model AI collaboration supporting OpenAI, Anthropic, and Google
- GitHub Integration for automated PR and Issue processing
- 30 minutes of free AI model usage per month
- Long-running task execution with real-time monitoring
- Custom Knowledge with RAG and Vector Store support
- Unlimited Document Vector Stores on free self-hosted plan
- Pre-built workflows like Deep Researcher and PRD Generator
- ISO/IEC 27001 compliance by default
- Separate build and execution environments for reliability
- Supports chain-of-thought agents for complex tasks
- Auto-selects best AI model for each task
- Email support on Pro plan
- Adapts to new AI models without rewrites
- Perfect for lean AI-native startup teams
Giselle cons
- Free cloud plan limited to 30 minutes AI usage per month
- Team Plan not yet available (coming soon)
- Free plan limited to single-user account only
- Pro plan unlimited usage is a limited-time promotional offer
- Additional usage costs 10% markup after credits on Pro plan
- SOC 2 certification still in progress
- GitHub Vector Stores limited to 3 on Pro plan
- No phone support available
- Downgrades take effect at end of billing cycle
- Free plan unused credits discarded when upgrading
Frequently asked questions about Giselle
What is Giselle?
Giselle is a platform that enables anyone to intuitively build AI applications with drag-and-drop operations, requiring no programming knowledge. It is designed as an execution platform to reliably complete complex, long-running AI tasks as business operations for engineers, designers, product managers, marketers, sales, and back-office professionals.
Do I need coding experience to use Giselle?
No. Giselle's visual, node-based interface is designed for user accessibility with drag-and-drop functionality. Users of all technical backgrounds can construct effective AI agents, making AI accessible across your team without writing code or tuning prompts.
What AI models does Giselle support?
Giselle supports multi-model AI collaboration, freely combining world-leading AI models like OpenAI GPT-5.2, Anthropic Claude Opus 4.5, and Google Gemini 3 Flash within a single workflow. You can assign roles that leverage each model's strengths for consistent results on complex tasks.
How does GitHub Integration work?
GitHub Integration automatically triggers AI apps based on GitHub events like Issues, Pull Requests, and comments. It enables context-aware processing that understands PR and Issue intent for code reviews, documentation generation, issue classification, and automatically updates READMEs, release notes, and blog content when code changes are merged.
What is the Visual Node Editor?
The Visual Node Editor lets you design AI workflows by placing and connecting nodes with drag-and-drop, just like drawing a flowchart. You can build complex AI processes visually without any programming experience by dragging nodes from the toolbar onto the canvas and connecting Output ports to Input ports.
How does Custom Knowledge work?
Custom Knowledge lets you incorporate your own data—PDFs, documents, GitHub repository code, Issues, and PRs—as knowledge for your AI apps. This achieves answers based on organization-specific information that general AI cannot provide, leveraging advanced RAG and Vector Store technology without specialized knowledge.
What is the difference between cloud and self-hosted deployment?
Cloud deployment is managed by Giselle for immediate use with a free plan offering 30 minutes of AI usage monthly. Self-hosted deployment uses the open-source version from GitHub with unlimited features, no usage limits, but requires your own infrastructure. Both have the same core features.
What are Vector Stores in Giselle?
Vector Stores enhance agent intelligence with your own data through Document Vector Stores (for PDFs and documents) and GitHub Vector Stores (for repository code, Issues, PRs). Free self-hosted has unlimited Document Vector Stores. Pro plan includes 5 Document and 3 GitHub Vector Stores. Team plan includes 20 Document and 10 GitHub Vector Stores.
How do I delete my account?
To delete your account, you must contact the Giselle support team at [email protected]. Account deletion is not self-service through the interface.
Is my data used for AI model training?
No. When you use Giselle, your data is sent to AI providers (Anthropic, OpenAI, Google) via their APIs. All these providers explicitly state that API data is not used for model training by default. Your prompts, inputs, and outputs are processed to generate responses but are not used to train or improve the AI models.