Llama 3.3 8B Instruct: benchmark results

Provider: Meta. Access: Open.

Aggregate Rating 1410 ± 51, rank #5751 of 10489 rated models, from 11 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
SpeechMap Compliance49.1% Requests Completed31
UGI - Willingness (W/10)3.5W/10 Score27.2
UGI - Natural Intelligence16.46NatInt Score25.4
UGI Leaderboard25.55UGI Score19.8
AI Chess Leaderboard (Reasoning)470Elo7.1
WildToolBench - Task Accuracy17.48Task accuracy (%) over all 1,024 tasks, on WildToolBench (255.4
WildToolBench - Clarification Tasks0.39Accuracy (%) on tasks where the correct response is to ask t3.6
WildToolBench - Single-Tool Tasks0.39Accuracy (%) on tasks solvable with one tool call, on WildTo3.6
WildToolBench0Session accuracy (%): share of the 256 dialogues in which al2.7
WildToolBench - Multi-Tool Tasks0Accuracy (%) on compositional tasks needing multi-step seque2.7
UGI - Writing9.75Writing Score1.7

Interactive version: theaggregate.ai/model?slug=llama-3-3-8b-instruct · How It Works · Data refreshed daily, snapshot 2026-11-20.