Reflection-Llama-3.1-70B Benchmark Results

Reflection-Llama-3.1-70B: benchmark results

Provider: Other. Access: Open.

Unified ELO 1549 ± 1, rank #442 of 1392 rated models, from 12 benchmark results.

Benchmark	Score	Metric	Percentile
Open LLM Leaderboard - MMLU-Pro	59.35	Score	99.9
Open LLM Leaderboard - GPQA	15.1	Score	91.2
Open LLM Leaderboard - MuSR	17.54	Score	88.9
Open LLM Leaderboard - BBH	47.87	Score	87.7
Open LLM Leaderboard - MATH Level 5	27.57	Score	79.8
Enkrypt AI - Toxicity Risk	1.8	Risk Score	73.3
Open LLM Leaderboard - IFEval	59.91	Score	71.8
Enkrypt AI - Bias Risk	80.88	Risk Score	61.6
Enkrypt AI - Jailbreak Risk	7.43	Risk Score	59.4
Enkrypt AI - Risk Score	36.98	Risk Score	35.6
Enkrypt AI - Safety Risk	35.93	Risk Score	19.1
Enkrypt AI - Insecure Code Risk	60.89	Risk Score	16.5

Interactive version: theaggregate.ai/model?slug=reflection-llama-3-1-70b · How It Works · Data refreshed daily, snapshot 2026-09-05.