Leaderboard Data

Leaderboard Data Directory

Where the raw data behind 40 LLM benchmark leaderboards lives, what its licence lets you do with it, whether you can submit results, and when it last changed. Every licence and submission statement is quoted from the operator's own page or repository and linked. Dates come from the last commit or dataset update.

Tap a box to filter by it; tap it again to remove that filter. Filters combine.

40 of 40 leaderboards. Where each leaderboard's data lives and what its licence says. Not legal advice; no scores are copied here.

LeaderboardRaw dataLicence / reuseSubmissionsLast updated
Aider LLM Leaderboards (polyglot)
Aider (Paul Gauthier)
GitHubattributionNot stated2025-11-20
345 days old
AlpacaEval
Stanford Tatsu Lab
GitHubattributionOpen2025-08-09
448 days old
ARC Prize leaderboard (ARC-AGI)
ARC Prize Foundation
not publishedno licenceRestrictednot published
Artificial Analysis
Artificial Analysis
Download / APIrestrictedNot statednot published
BenchLeader
BenchLeader
Download / APIattributionNot statednot published
BenchLM
BenchLM.ai
Download / APInon-commercialNot statednot published
Berkeley Function Calling Leaderboard (BFCL)
UC Berkeley (Gorilla)
GitHubunclearNot stated2026-04-12
202 days old
BigCodeBench
BigCode project
Hugging FaceunclearClosed2025-10-15
Archived
Epoch AI Benchmarking Hub
Epoch AI
Download / APIattributionNot stated2026-10-31
EQ-Bench
Samuel Paech
not publishedunclearRestricted2026-05-31
153 days old
EvalPlus (HumanEval+ / MBPP+)
EvalPlus team
not publishedunclearNot stated2025-10-02
394 days old
GAIA
GAIA team (Meta, Hugging Face)
Hugging Faceno licenceNot stated2026-10-31
HELM
Stanford CRFM
Download / APIunclearNot stated2026-06-05
148 days old
Humanity's Last Exam (HLE)
Center for AI Safety and Scale AI
not publishedunclearNot stated2026-09-22
K2 Vendor Verifier
Moonshot AI
GitHubno licenceNot stated2026-02-14
259 days old
Kagi LLM Benchmarking Project
Kagi Inc.
not publishedno licenceNot stated2026-02-13
260 days old
LiveBench
LiveBench team / Abacus AI
Hugging FaceunclearNot stated2026-10-25
LiveCodeBench
LiveCodeBench team
not publishedunclearNot stated2025-07-16
472 days old
LLM Confabulation Leaderboard
Lech Mazur
GitHubno licenceNot stated2025-08-07
450 days old
LLM Stats
llm-stats.com
not publishedattributionNot statednot published
LMArena (Arena) leaderboard
LMArena
Hugging FaceattributionNot stated2026-10-30
MathArena
SRI Lab, ETH Zurich
Hugging FaceunclearNot stated2026-10-19
MMMU / MMMU-Pro
MMMU team
not publishedunclearNot stated2025-09-05
421 days old
MTEB Leaderboard
MTEB (embeddings-benchmark)
GitHubHugging FaceCC0Open2026-10-30
NYT Connections (Extended)
Lech Mazur
GitHubno licenceNot stated2026-10-23
Open LLM Leaderboard
Hugging Face
Hugging Faceno licenceClosed2025-03-20
Retired
OpenRouter rankings
OpenRouter, Inc.
not publishedno licenceNot stated2026-10-30
OpenVLM Leaderboard (OpenCompass)
OpenCompass (Shanghai AI Lab)
Download / APIunclearNot stated2026-03-11
234 days old
OSWorld / OSWorld-Verified
XLANG Lab (HKU)
not publishedunclearNot stated2026-10-24
QuantDiff
QuantDiff
Download / APIattributionNot stated2026-10-30
RewardBench / RewardBench 2
Allen Institute for AI (Ai2)
Hugging FaceunclearNot stated2025-12-11
324 days old
Scale SEAL Leaderboards
Scale AI
not publishedno licenceRestrictednot published
SimpleBench
SimpleBench Team
not publishedunclearNot stated2024-12-20
680 days old
SWE-bench
SWE-bench team (Princeton / Stanford)
GitHubno licenceRestricted2026-09-01
Terminal-Bench
Stanford / Laude Institute (Harbor)
Hugging FaceattributionClosed2026-05-15
169 days old
UGI Leaderboard
DontPlanToEnd
not publishedunclearNot stated2026-10-17
Vals AI benchmarks
Vals AI
not publishedno licenceNot stated2026-10-29
Vectara Hallucination Leaderboard
Vectara
GitHubattributionNot stated2026-10-23
WildBench
Allen Institute for AI (Ai2)
not publishedunclearNot stated2024-08-01
821 days old
τ-bench (tau2-bench)
Sierra
GitHubattributionOpen2026-10-30

At a glance

How the labels work and what is left out: Methodology. Changes to this list: Changelog.

Leaderboard changes in your inbox

An occasional short email when leaderboards change their licence or submission rules, publish new raw data, retire, or are added here. No spam. Every email has a one-click unsubscribe link.