I tried every AI dev tool so you don't have to

AI Coding Tools: My Top Picks

Overview of AI Coding Tools

• The speaker spent three months testing various AI coding tools, ultimately canceling subscriptions to all but a select few that proved effective.

• The chosen tools are not the most popular or hyped but are those the speaker uses daily for engineering and research tasks.

Models: The Brains Behind Applications

• GPT 5.5 is highlighted as a top choice for complex reasoning and architectural work, marking a significant improvement over previous models.

• Opus 4.7 excels in generating UI elements, with the speaker using it to create visually appealing components by referencing designs from Dribbble.

• Gemma 4 is praised for its local coding capabilities, particularly useful for parsing transcripts and creating content post-video recording.

• Gemini 3 Pro is utilized for deep research and feature planning, leveraging URL context to enhance search capabilities.

• Sonnet 4.5, despite being considered outdated, is favored for its low token consumption and speed in handling simpler tasks without needing advanced intelligence.

Essential Tools: Authentication and Project Setup

• The speaker emphasizes the importance of Clerk for authentication in projects, especially after the introduction of Clerk CLI which streamlines project setup significantly.

• Clerk CLI allows developers to configure user flows directly from the terminal, enhancing efficiency compared to traditional dashboard methods.

Harnesses: Enhancing Workflow Efficiency

• The integration of Plaud Code with Cursor creates an unmatched workflow experience by providing contextual assistance across codebases.

• Cursor's auto-completion features are highlighted as essential for writing complex code without needing to memorize every detail.

• Multiple agents can run simultaneously within Cloud Code, allowing tailored task management while maintaining productivity.

Connecting Everything: Subscriptions and Voice-to-Code Features

• Key subscriptions include Cloud Code max plan and Gemini Pro; these provide necessary resources without hitting usage limits during intensive work periods.

• Whisper Flow AI stands out as a voice-to-code tool that enhances productivity by converting spoken commands into code efficiently while on-the-go.

Final Thoughts on Tool Selection

• The speaker encourages viewers to avoid unnecessary spending on multiple similar tools; instead, focus on finding models that fit specific workflows effectively.

• Budget considerations should guide model selection; older models like Sonnet 4.5 can still deliver satisfactory results depending on task requirements.