4 ms·
> Anthropic is claiming that Claude 5 Sonnet will cost about half as much as their current SOTA models. Therefore, expect about half the performance. That's no
by minimaxir 9mo ago
> Anthropic is claiming that Claude 5 Sonnet will cost about half as much as their current SOTA models. Therefore, expect about half the performance.
That's not how LLM quality works.
- cadabrabra 9mo agoMaybe not in theory but definitely in practice, as we’ve seen with GPT-5. These companies are lightning money on fire. If they reduce the cost, expect a proportional decrease in quality. All of the GPT-5 anecdotes confirm this. When the data and anecdotes disagree, the anecdotes are usually right, and the data is usually bullshit.
- minimaxir 9mo agoGPT-5's issues were due to router shenanigans which Claude models do not do.
- cadabrabra 9mo agoNo dude, the latest versions of the models it routes to are markedly poorer in performance than their predecessors. I’m observing a law that states: There appears to be a direct relationship between model performance and cost, such that whenever a company claims to have reduced inference costs, customers immediately notice a corresponding decline in model performance.