came back to lovable from claude code last week. heres why, in case youre considering the same.
ok so this is the inverse of every "i migrated off lovable" post you keep seeing. in case youre also lost on which tool is correct.
quick history. moved my main project from lovable to claude code in november. genuinely had real reasons at the time. credits got expensive, i wanted more control over backend logic, claude code is just more powerful for complex stuff.
5 months in claude code. ship velocity dropped maybe 60%.
yes the agent is better at deep work. no the agent is not better at "lets add a pricing card and link it." that was 3 prompts in lovable. that was 18 prompts in claude code, because claude code wanted to discuss the right abstraction for a pricing card component and i wanted to ship.
then opus 4.7 landed in lovable on apr 24. i tried it on a new feature for the same project. just to see.
it was so much faster than what i had been doing.
Like, embarrassingly. The thing that took me a full saturday in claude code took me 90 minutes in lovable.
now im running both. claude code for backend stuff that needs careful thinking. lovable for everything else. probably 80/20 lovable.
what i think most people get wrong about the migration question:
claude code is not "the upgrade." its a different tool for different work
lovables ceiling is higher than 6 months ago because of plan mode + workspace knowledge + opus 4.7
if your project is "saas with a normal-shaped frontend" lovable is probably correct
if your project is "weird backend with custom infra needs" claude code is probably correct
if your project is both, dont pick. use both.
would i make the original migration again? probably not. i overweighted credit cost and underweighted speed cost. credit cost shows up on a dashboard. speed cost shows up in the form of "its been 4 hours and i havent shipped anything."
anyone else come back? curious if im the only one doing this or if its a quiet trend.