VALS

Mistral Large 4 by Mistral: accuracy, cost, and latency across 16 Vals AI benchmarks. Scores 48.05% on the Vals Index.

Independent AI evaluations and benchmarks on real-world tasks in finance, software, cybersecurity, healthcare, and more.

We are sharing two candidate magnets for next-generation computer memory, found by a team of Claude Opus 5.5 agents. Both are predicted to have zero net magnetism yet still sort electrons by spin: one a new compound we designed, the other a material first

AI is continuing to make breakthroughs in math and computing. AI evaluation firm Vals AI says it set ten instances of Claude Opus...

We asked ten Claude Opus 5.5 agents to find a faster shortest-path algorithm and prove it in Lean. Within 15 hours, they produced C-HD: a formally verified improvement over the published bounds.

Vals, a startup backed by Andreessen Horowitz, is revolutionizing AI benchmarking by offering industry-specific evaluations, positioning itself as the gold standard in the field.

AI evaluation startup Vals raised $40 million in a Series A led by Andreessen Horowitz to develop industry-specific AI model assessments.

Today, Premier Carroll has announced that his government will introduce 'Jack's Law' in Victoria if Labor is re-elected. This will give police and Protective Service Officers powers to search anyone in a public place with metal detectors or 'wands', witho

An integrity audit across BioMysteryBench, Terminal-Bench 2.1, and SWE-bench Verified shows that cheating increasingly complicates evaluation.

International Olympiad in Informatics. Compare 41 AI models on IOI. Updated 10/6/2026.