|
|
|
|
|
|
by hnlmorg
3 days ago
|
|
|
To be honest even the cloud models are a hot mess at times. This week I’ve spent more time rejected code from OpenAI models than I have approving it. In fact it really feels like OpenAI models have taken a nose dive this week compared with Claude. At least for my specific workloads (these things are so variable it’s like trying to compare Google results…) |
|
|