Free Cursor Alternative? Trae AI Just Got Way Better

Key Concepts

• AI Coding Assistant: Trey AI as a free alternative to other coding IDEs.
• Custom Agents: Customizable coding agents for specific tasks with specific tools.
• MCPS (Modular Code Processing System): Support for MCPS, enabling integration with various tools and APIs.
• Web Search Integration: Ability to perform web searches directly within the IDE.
• Model Selection: Support for various models like Cloud 3.5 Sonnet, Gemini 2.5 Pro, and custom models via Open Router.
• Privacy and Security: Local data storage, secure data access, and regional deployment for data regulation compliance.

Trey AI: A Free AI Coding IDE

Trey AI is presented as a free AI coding assistant and a full-fledged coding IDE built on top of VS Code. It's highlighted as a good alternative to other coding IDEs, especially with its recent upgrade that includes customizable coding agents and MCPS support.

New Features and UI Redesign

Trey AI has undergone a UI redesign and introduced new features:

• Custom Agents: Users can create and customize agents for specific tasks.
• MCPS Support: Integration with Modular Code Processing Systems for enhanced functionality.
• Custom Rules: Ability to define custom rules for projects.
• Web Search: Integrated web search capabilities.

Custom Agents and MCPS Integration: A Detailed Walkthrough

The video focuses on the custom agent feature as a standout capability. The process of creating and configuring a custom agent is demonstrated:

1. Creating a New Agent: Naming the agent (e.g., "Web Search").
2. Customizing Agent Behavior: Providing a custom prompt to define the agent's behavior.
3. Selecting Tools: Choosing which tools the agent has access to (e.g., preview, web search, terminal, file system).
4. Adding MCPS: Integrating MCPS for specific functionalities, such as web search using the Tavi API.
5. Configuring MCPS: Adding necessary configurations, including API keys.
6. Enabling the MCP: Activating the MCP within the custom agent.

Example: Web Search Agent with Tavi MCP

The video provides a step-by-step example of creating a "Web Search" agent using the Tavi MCP for web search functionality:

• Tavi API: An API that enables web search capabilities.
• Configuration: Copying the required configuration from the Tavi documentation.
• API Key: Providing the Tavi API key.
• Testing: Asking the agent to look up the latest Gemini SDK to test the web search functionality.

Project Example: Text-to-Image Generator

The video demonstrates how Trey AI can be used for a real-world project: a client-side web application for generating images from text prompts using the Google Gemini API.

Requirements:

• Use the Gemini API to generate images based on text prompts.
• Utilize the Flash 2.0 model for native image generation.
• Support image regeneration and downloading.
• Display an animation while waiting for image generation.

Implementation:

1. Providing Requirements: Using the pound sign (#) to provide the requirements document to the builder agent.
2. Code Implementation: The agent implements the project based on the requirements.
3. API Key: Providing the Gemini API key.
4. Code Execution: Running the Python code in the terminal within Trey AI.
5. Web Page Preview: Previewing the web page within the IDE.
6. Image Generation: Testing the image generation functionality with a prompt (e.g., "a llama playing soccer").
7. Debugging: Addressing errors by providing screenshots and specific instructions to the agent.
8. Functionality Testing: Verifying image regeneration and download functionalities.

Coding Capabilities and UI

• Code Explanation: Ability to select code and ask the AI to explain it.
• Model Dependency: Coding capabilities depend on the chosen model (Gemini 2.5 Pro, Cloud 3.5 Sonnet, GPT-4.1).
• Code Editing: Editing code directly within the IDE.
• Diff View: Clear diff view for code changes.
• Clean UI: The UI is highlighted as clean compared to other coding IDEs.

Privacy and Security Considerations

Trey AI's approach to privacy and security is addressed:

• Local Data Storage: Codebase files are stored locally on the user's device.
• Secure Data Access: Strict access control and encrypted transmissions.
• Regional Deployment: User data and infrastructure are deployed based on account location (US, Singapore, Malaysia) with isolation to meet local data regulations.

Quote: "Local data storage codebase files are restored locally on your devices for indexing files may be temporarily uploaded for embedding after processing all plain text is deleted." - Trey AI Website

Synthesis/Conclusion

Trey AI is presented as a powerful and free coding IDE with unique features like customizable agents and MCPS support. It offers a clean UI, supports various coding models, and provides tools for efficient coding and debugging. While privacy and security are addressed, users are encouraged to evaluate it and compare it with other coding IDEs like Cursor and Windsurf. The ability to create custom agents for specific tasks, as demonstrated with the web search and text-to-image generation examples, is a key differentiator.