IMHO this is a difficult question to answer. Part of the power of paid models comes from the software supporting it. With local models, you have tons of workflows that can severely influence the quality of the result.
In my personal experience, the SOTA models are way more consistent and can handle more complex questions. Part of that is (probably) because I don't let my local model access the internet, while paid models do use the internet to look at docs etc.
Reading the actual code is also always a better source of truth than docs anyway (this is true for people and LLMs). Just clone whatever libraries you use.