โ† sources

How we all feeling about the new Anthropic model updates? ๐Ÿซก

source: reddit (reddit:ClaudeCode)published 2026-11-10 17:26ingested 2026-11-11 02:42

retained text

I haven't had the chance to fire up the new Anthropic models myself yet, but looking at the benchmarks making the rounds, it feels like Anthropic is forcefully taking the crown back.

Between the massive leaps in coding proficiency and the agentic benchmarks everyone is screaming about, the data sheets look like pure wizardry. But we all know synthetic benchmarks and actual, messy production code are two very different beasts.

For those of you who have already plugged the new models into your terminal or claude-code workflow:

. How does the speed and latency actually feel when it's actively refactoring? 

. Is it actually one-shotting complex bugs, or is it just confidently hallucinating at a higher velocity?

. Does it feel like a genuine generational leap, or is it just incremental polish wrapped in good PR?

Drop your raw, unvarnished verdicts below. Save me from relying on corporate graphs!

document fetched 2026-11-11 00:40 ยท canonical https://reddit.com/r/ClaudeCode/comments/1xssxe8/how_we_all_feeling_about_the_new_anthropic_model