Claude Opus 5.5 Takes Top Spot on Vals Index

Vals AI ranks Claude Opus 5.5 first on its Vals Index, an independent benchmark of coding and agentic performance, up two spots and two points from Claude Opus 5. That puts Anthropic in all three of the index's top positions. Vals AI notes GPT-6 Sol results weren't yet available on the index at the time of this post.

The two-point, two-spot move is a modest gain on this specific index rather than a dramatic reordering, and it's one data point among several benchmarks published that same week: Cursor's own coding benchmark and Perplexity's WANDR benchmark separately rated Opus 5.5 well on cost-adjusted performance, though each measures a different task set and shouldn't be read as confirming this specific ranking.