AI Models & LLM News: Releases, Benchmarks, Comparisons
New models, benchmarks and our tests
Opus 4.7 is worse than Opus 4.6 at spotting fakes
Opus 4.7: 50% More Tokens, 5h Limits, Devs Are Mixed
Claude Opus 4.7 Hits 87.6% on Agentic Coding, Beats 4.6
Opus 4 to 4.7: the quiet shift no one’s pointing at