openbench provides standardized, reproducible benchmarking for LLMs across 30+ evaluation suites (and growing) spanning knowledge, math, reasoning, coding, science, reading comprehension, health, long ...
This project has been developed to provide a shared conditional logic between front-end and back-end code, stored in JSON or in any other data serialization format ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results