Everyone is rushing to route prompts through different LLMs, but it is rarely a...

https://charlie-wiki.win/index.php/Relay_Workflows_with_Multiple_Models:_Who_Goes_First%3F

Everyone is rushing to route prompts through different LLMs, but it is rarely a silver bullet. Mixing models helps you tap into varied reasoning strengths and reveals where they disagree, which is great for mapping failure modes. But proceed with caution