In just 3 weeks, we created an open source replacement for Manus! Contribute source code and use it for free

In addition, Suna integrates a powerful toolset, including browser automation for web browsing and data extraction, file management capabilities for document creation and editing, web crawling and enhanced search capabilities, website deployment capabilities, and the ability to integrate with multiple APIs and services.

These features work together to enable Suna to solve complex problems through simple conversations and automate various workflows!

Since it’s free, Synced also tried it out. The prompt was: “Help me compile a list of coffee brands from around the world.”

Suna first mapped it out and organized it by country or region.

Then, based on the to-do list, Suna uses web search to collect comprehensive information about coffee brands in different regions, during which it searches various web content.

Finally, Suna compiled all of this information into a comprehensive document about coffee brands around the world, breaking them down by region.

In addition to personal experience, the official also released a lot of demos.

First let it analyze the US stock market.

Users only need to enter a sentence: "Write me a detailed report on what happened in the US stock market over the past two weeks. Analyze the trend of the S&P 500 index and tell me what the market expects in the coming weeks. This report is written for the CFO of a bank."

Suna outputs while thinking. After about a minute of execution, a complete and readable report is produced:

Next, Suna's ability to analyze website content was tested.

Input prompt: "Visit Crunchbase, Dealroom, and TechCrunch, filter by Series A funding in the SaaS finance space, and build a report for outbound sales with company data, founders, and contact information."

Again, Suna did a great job.

Let’s look at another example: “Please help me research the B2C AI market and show me some interesting data charts you find. Please browse the web using your browser.”

After watching various demonstrations, Suna is really an excellent AI assistant.

At present, the project has just been launched not long ago, and the number of star labels has increased rapidly.

Project address: https://github.com/kortix-ai/suna

According to the project introduction, Suna consists of four main components:

1. Backend API : A Python/FastAPI-based service that handles REST endpoints, thread management, and integration with large language models (LLMs) such as OpenAI and Anthropic through LiteLLM.

2. Frontend : Application built with Next.js/React, providing responsive user interfaces including chat interface, dashboard, etc.

3. Agent Docker : An isolated execution environment for each agent with browser automation, code interpreter, file system access, tool integration, and security features.

4. Supabase database : responsible for data persistence, including user authentication, user management, conversation history, file storage, agent status, analysis, and real-time subscription functions.

Finally, according to public information, Kortix was founded in 2024 and its CEO is Marko O. Kraemer. The company focuses on developing AI Agents, aiming to complete complex tasks in the real world through natural dialogue.