"The community" is astroturfed as hell though. Anthropic pays influencers to promote Claude Code and likely bots a ton as well, so it's hard to come to any kind of consensus online. Even if everyone was acting in good faith, some people will have a much better experience than others because of the domain they're working in (e.g. AI being much better at frontend and commonly used libraries).
The only real way to evaluate a model is to test it yourself but that's exhausting for each new model and not comprehensive anyway.
The only real way to evaluate a model is to test it yourself but that's exhausting for each new model and not comprehensive anyway.