Sonnet 5.5 was just released yesterday, and there was one thing that shocked me!
Terminal-Bench 4.0 (agentic coding) Sonnet 5.5: 70.6% Opus 5.5: 66.4% So basically, Sonnet 5.5 beats Opus 5.5 in this benchmark. In several other benchmarks, Opus 5.5 still wins, but slightly
Even though it's the same song, the expression changes so much 😳 In Sonnet 5.5, which came out today
Even though it's the same song, the expression changes so much 😳 In Sonnet 5.5, which came out today, I created the lyric motion for "Ninety-Nine" made with Suno. The top is Sonnet 5.5 and the bottom is Opus 5.5. If you compare them, you'll notice that the way the characters appear, how they move, and the way they take the "pauses" on the screen are quite different. Rather than saying which is correct,…
Sorting race with 4 models including Jev AI [Simultaneous comparison with Opus 5, GPT-6 Astra, Gemini 3.8 Flash]
・Throw the same sorting to 4 models at once ・Gemini 3.8 Flash takes surprisingly long time ・The difference is obvious even with terminal commands What would you judge?
Opus 5 (not 5.5) seems to not be as bad at motion graphics as I thought.
Still better taste than sol and even astra imo.
You have reached the end of the archive
All of Claude Opus 5