About to pull the trigger on switching from Claude, back to OpenAI after 2 years.

It's just too much value at the table at this point. Anthropic has been too late on so many things: Multimodal, larger context windows, agentic use, empathetic models, etc...

Any final thoughts on why I shouldn't switch?

https://chainleft.art

Dad, data scientist, blockchain native artist.

Power consolidation=evil.

I try to like Claude and did genuinely for a while but struggle to get it to meet any of my needs now.

I definitely prefer local ollama models for a lot of things, mostly just use OpenAI when I need more firepower

compressionism
|
art guy /artonly
|
linktr.ee/0xdiid

Out of the box but purpose specific, gotta fine tune the prompts a LOT more to get what you want though.

Interesting. Are you local models finetuned / retrained or out of the box?