I keep seeing a lot of people saying X is the best AI for coding. My advice to every company still trying to decide, pick a few and let your developers try them out.
Personally, I've found that for my PERSONAL projects, Github's Copilot is the best bang for the buck.
I get access to a variety of frontier models I get to decide which model I like best I can code my way
The fun part is that if I like claude code's cli model of coding (I'm warming up to it) I can use opencode and connect it to my github copilot account.
Claude Code and the Claude LLMs iares good, don't get me wrong, but I've not found them to be the best. Maybe it's the problems I've been solving.
Experiment, try different models and different tools before committing to a single one.
Also, don't overlook openai's Codex. I was prepared to hate it but I'll have to say I was impressed.
Originally posted on Linkedin: Try Multiple AI Coding Tools Before Committing.