Hi HN — I’m building The Tank (thetank.io), a multi-model AI workspace designed to reduce the “AI tab chaos” problem.
The core idea:
- One place to chat, think, and build
- Multiple AI models working together (instead of switching tabs constantly)
- A builder workflow that aims to support app/site generation in the same environment
Why I’m building it:
I kept finding myself using one tool for brainstorming, another for coding, another for debugging, another for rewriting copy, etc. The context switching was brutal. I wanted a single workspace where a “collective” of models could help with the same task.
What I’m actively working on:
- Better builder/live preview experience
- More reliable code persistence across sessions
- Smoother transitions from chat -> build -> iterate
What I’d love feedback on:
1. Is the “multi-model collective” idea compelling, or do people prefer one model per task?
2. What would make this actually replace your current workflow?
3. If you’ve used v0/Bolt/Lovable/etc., what’s the most important UX behavior to get right?
I’d especially appreciate blunt feedback from people who build with AI every day.
