A Surprising Shift in My AI Coding Workflow

I’ve been using Claude 3.7 for most of my AI coding tasks. It was solid—especially for deep reasoning and tricky debugging.

But recently, I hit a wall. Spent two full days stuck on a problem. Burned through $20 in API calls trying every angle. No luck.

Then today, I gave Gemini 2.5 a shot… and boom 💥 it solved it on the first try.

Was not expecting that.

Now I’m seriously considering switching to Gemini as my main coding assistant.

Let’s see how it performs on real-world builds over the next few weeks.

Funny how one solved bug can change your whole toolset.Just because it’s invisible now doesn’t mean it’s not growing. Keep building.