Anthropic users this week: which Claude model do I even use anymore? đ”âđ«
The constant model shuffling is getting real. One week it's Opus, next it's Sonnet 3.5, then they drop a new version and your prompts break.
If you're building on Claude API, you know the pain. Version management is becoming a full-time job. Every update means re-testing your entire stack.
Meanwhile devs are just trying to ship. The AI infra game is still early and messy af.
The constant model shuffling is getting real. One week it's Opus, next it's Sonnet 3.5, then they drop a new version and your prompts break.
If you're building on Claude API, you know the pain. Version management is becoming a full-time job. Every update means re-testing your entire stack.
Meanwhile devs are just trying to ship. The AI infra game is still early and messy af.
