[__natty__]

You compare tiny modal for local inference vs propertiary, expensive frontier model. It would be more fair to compare against similar priced model or tiny frontier models like haiku, flash or gpt nano.

[javawizard]

Not when the article they're commenting on was doing literally exactly the same thing.

[ericd]

Eh it’s important perspective, lest someone start thinking they can drop $5k on a laptop and be free of Anthropic/OpenAI. Expensive lesson.