Llama 3 70B Instruct

Llama / Meta

29scores
25benchmarks
$0.51 / $0.74 per 1M tokenscost in/out

Metadata

Llama Open source

Aliases: llama-3-70b-instruct, meta-llama-llama-3-70b-instruct, meta-llama/llama-3-70b-instruct

Benchmark Results

Benchmark Category Rank Score Sampled
RewardBench Alignment 79 76.27 2026-05-06
BigCodeBench Coding 37 43.60 2026-05-06
ENAMEL Coding 5 0.42 2026-05-06
EvalPlus Coding 20 70.50 2026-05-05
HumanEval+ Coding 24 72 2026-05-05
MBPP+ Coding 19 69 2026-05-05
NeoEvalPlusN Creative 41 16.75 2026-05-06
Open FinLLM Leaderboard Finance 7 31.228277% 2026-05-27
MixEval Chat General Knowledge 12 55.90 2026-05-06
Open LLM Leaderboard v2 General Knowledge 534 36.37 2026-05-06
CyberSecEval Generalization 3 29.27% 2026-05-27
L-Eval Generalization 2 68.355% 2026-05-27
AgentClinic Healthcare 6 19.0% 2026-05-27
HealthBench Hard Healthcare 19 0.41 2026-05-27
MuSR Intelligence 2005 10.92 2026-05-06
AraGen v3 Language 47 26.14 2026-05-06
Open Japanese LLM Leaderboard Language 161 62.57 2026-05-06
Open Japanese LLM Leaderboard Language 437 51.07 2026-05-06
Open Portuguese LLM Leaderboard Language 26 86.73 2026-05-06
MATH Level 5 Math 1018 24.47 2026-05-06
OlympiadBench Math 7 20.27 2026-05-06
MEDIC Benchmark Medical 61 60.75 average normalized public table score 2026-05-27
BenchBench Meta 19 0.82 2026-05-06
LanguageBench Multilingual 18 0.51 2026-05-06
ChemBench Science 24 0.52 2026-05-06
ChemBench Science 25 0.52 2026-05-06
VNTL Leaderboard Translation 39 66.91 2026-05-06
VNTL Leaderboard Translation 39 65.94 2026-05-06
VNTL Leaderboard Translation 54 63.30 2026-05-06