Ruli published its results from the 2026 Legal Benchmarks evaluation, a test designed by lawyers that uses real in house contract and extraction tasks. It reports an overall score of 72.0 against a legal AI average of 63.7, a full task pass rate of 31.1% against 24.7%, and 41.7% of data extraction tasks fully correct against 31.3%.
Buyer relevance. Scores on a shared benchmark are more useful than case studies, because buyers can compare them across vendors. A full task pass rate near one in three is also a reminder that most outputs still need a lawyer's correction.