Suna: FULLY FREE Manus Alternative with UI! Generalist AI Agent! (Opensource)

Key Concepts:

• Sunno AI: An open-source, all-in-one AI agent for tasks like research, data analysis, and workflow automation.
• Local AI Agent: Running the AI agent on your own computer instead of relying on cloud services.
• Open-Source: Software with publicly accessible source code, allowing for modification and distribution.
• Backend API: Python Fast API handling endpoints, thread management, and LLM integrations.
• Frontend: Next.js and React-based user interface.
• Agent Docker: Isolated Docker environment with browser automation and code interpreter.
• Superbase: Database for authentication, user management, and conversation history.
• LLM (Large Language Model): The AI model used for natural language processing (e.g., Anthropic, OpenAI, Gemini).
• Docker Compose: A tool for defining and running multi-container Docker applications.
• Tavily API: Optional API for enhanced search capabilities.
• Firecrawl: Optional API for web scraping capabilities.

1. Introduction to Sunno AI

• Sunno AI is presented as a compelling open-source alternative to Manis and other advanced AI agents.
• It's highlighted as being free, fully local, and accessible to everyone.
• Sunno AI is designed for real-world productivity, enabling users to tackle tasks like research, data analysis, and daily workflows through natural language.
• The key focus is on delivering the results that users specifically request.

2. Sunno AI Architecture and Components

• Sunno AI's functionality relies on four key components:
  • Backend API: A Python Fast API responsible for powering the rest endpoints, managing threads, and integrating with large language models (LLMs).
  • Frontend: Built with Next.js and React, providing the user interface.
  • Agent Docker: An isolated Docker environment equipped with browser automation, a code interpreter, and other tools.
  • Superbase: A database used for authentication, user management, and storing conversation history.

3. Example Use Cases and Capabilities

• Market Analysis Example:
  • A prompt is given to Sunno AI to analyze the market for a healthcare company in the UK, including identifying major players, their market size, strengths, and weaknesses (SWOT analysis).
  • The agent successfully scours the web, utilizes a built-in web browser agent to analyze websites, and generates a PDF report of the market analysis.
• YouTube Channel Analysis Example:
  • The presenter used Sunno AI to create a dashboard analyzing their YouTube channel ("World of AI").
  • By providing the channel link, Sunno AI scraped the content and generated a dashboard with accurate data, including the number of videos and various charts.
  • This was achieved with a single prompt, showcasing the impressive capabilities of the open-source AI agent.

4. Setting Up Sunno AI Locally

• Prerequisites:
  • A Superbase project (free to set up).
  • Superbase CLI installed on the computer.
  • Optional: Redis for caching, Daytona for secure agent execution, Tavily API for enhanced search, and Firecrawl for web scraping.
  • Python 3.11 for the API backend.
  • Git installed.
  • An IDE like VS Code.
• Installation Methods:
  • Docker Compose (Recommended): The presenter uses Docker Compose for a simpler setup process. Docker Desktop needs to be installed.
  • Manual Setup: Involves setting up the Superbase project, Redis, Daytona, Python, and LLM provider separately.
• Step-by-Step Installation Process (using WSL on Windows):
  1. Clone the Sunno AI GitHub repository using WSL.
  2. Navigate to the Sunno AI directory using the command line.
  3. Configure the backend by navigating to the backend directory and creating an .env file.
  4. Fill in the required variables in the .env file, including Superbase credentials, LLM provider details, and API keys for optional services like Tavily and Firecrawl.
  5. Set up the Superbase CLI by logging in and linking the project.
  6. Push the database migration using the Superbase CLI.
  7. Configure the frontend by navigating to the frontend directory and creating an .env.local file.
  8. Fill in the required variables in the .env.local file, such as the Superbase URL, anonymous key, and backend URL.
  9. Install dependencies using either Docker Compose or manual commands (npm install in the frontend, then installing backend requirements).
  10. Start the application using Docker Compose or by running the frontend and backend servers simultaneously.

5. Demonstration of Local Sunno AI Usage

• After successful installation, Sunno AI can be accessed through a local host URL.
• The presenter demonstrates its usage by prompting it to perform a market analysis on Tesla stock and create a dashboard showcasing the findings.
• Sunno AI creates a plan, utilizes the Tavily API for web searches, and employs Firecrawl and the browser agent to scrape content.
• The resulting Tesla stock analysis dashboard includes insightful data, stock graphs, key insights, and technical outlooks.

6. Conclusion

• Sunno AI is presented as a competent, open-source alternative to Manis with a user-friendly UI.
• The video provides a detailed guide on setting up and using Sunno AI locally.
• The presenter encourages viewers to explore Sunno AI and suggests the possibility of future videos showcasing its capabilities.