Local AI Registry
COLLECTION / BENCHMARK

Leaderboards

Scraped public quality scores from GitHub Pages leaderboards such as Terminal-Bench 2.1. These are not local speed measurements.

24 records
AGIEvalagievalreasoningcategory491scored models89.2top scorereasoning
ARC-Challengearcknowledgecategory98scored models97.4top scoreknowledge
BrowseCompbrowsecompagenticcategory78scored models100top scoreagentic
GPQAgpqareasoningcategory421scored models89top scorereasoning
GSM8Kgsm8kmathcategory439scored models96.4top scoremath
HellaSwaghellaswagknowledgecategory392scored models91.8top scoreknowledge
HumanEvalhumanevalcodingcategory429scored models94.1top scorecoding
Humanity's Last Examhleknowledgecategory330scored models54.7top scoreknowledge
LiveCodeBenchlivecodebenchcodingcategory476scored models89top scorecoding
LongBenchlongbenchlong-contextcategory25scored models61.9top scorelong-context
MATHmathmathcategory1015scored models99.6top scoremath
MBPPmbppcodingcategory323scored models90.2top scorecoding
MGSMmgsmmathcategory143scored models92top scoremath
MMLUmmluknowledgecategory1152scored models94.4top scoreknowledge
MMLU-Prommlu-proknowledgecategory465scored models88top scoreknowledge
OpenBookQAopenbookqaknowledgecategory81scored models89.8top scoreknowledge
PIQApiqaknowledgecategory204scored models93.3top scoreknowledge
RULERrulerlong-contextcategory109scored models96.8top scorelong-context
SciCodescicodecodingcategory47scored models100top scorecoding
SWE-bench Verifiedswebench-verifiedswecategory61scored models77.8top scoreswe
Terminal-Benchterminal-benchagentic-codingcategory119scored models47.1top scoreagentic-coding
Terminal-Bench 2.1terminal-bench-2.1agentic-codingcategory17scored models82.7top scoreagentic-coding
WinoGrandewinograndeknowledgecategory297scored models89.5top scoreknowledge