Llama 2 70B Chat HF — Benchmarks

Benchmark scores for Llama 2 70B Chat HF aggregated from public leaderboards (9 boards), with how it ranks among open models and how it compares with its nearest neighbours. See hardware requirements for what you need to run it.

llmrun Composite Score →

50.4

90% range 31.8–69.0 · from 4 boards

CodingNot enough data yet
AgentsNot enough data yet
MathNot enough data yet
ScienceNot enough data yet
Reasoning45.390% range 21–69

Llama 2 70B Chat HF reasoning benchmarks

Multi-step logic and problem-solving — can the model think through a hard problem rather than recall an answer.

GPQA Diamond
GPQA Diamond is a set of extremely hard, graduate-level science questions (physics, chemistry, biology) written by domain experts and filtered so that skilled non-experts with web access still fail. It measures genuine reasoning rather than memorization.
26.3%
26.3%
median 54.0%open leader 93.1%
Overall best: Gemini 4 Argon (high) 96.4%
10th percentile · #88 of 98 open · #288 of 300 all
#88 of 98
#288 of 300
Independent
Epoch AI · Jul 2023
DTBench
DTBench is a set of handcrafted, expert-validated multiple-choice questions on decision theory — Newcomb-like problems and related puzzles — built by the Conceptual Reasoning Index project with Anthropic. It tests whether a model can follow an abstract argument rather than recall a fact.
41.6%
41.6%
median 67.7%open leader 93.9%
Overall best: Claude Opus 5.5 Max 98.9%
lowest · #68 of 68 open · #214 of 214 all
#68 of 68
#214 of 214
Lab-reported
Epoch AI · Jul 2023
ForecastBench
ForecastBench, from the Forecasting Research Institute, asks models to predict real-world future events, with questions that only resolve after the model's training cutoff so the answers can't have been memorised. It is scored as a difficulty-adjusted Brier index, where higher is better.
51.4
51.4
median 58.1open leader 61.9
Overall best: O3 (Apr 16, 2025, unspecified) 62.5
lowest · #26 of 26 open · #81 of 82 all
#26 of 26
#81 of 82
Lab-reported
Epoch AI · Jul 2023

Other results (1)· leaderboard-only, not part of the llmrun Score

BIG-Bench Hard
BIG-Bench Hard (BBH) is the 23-task subset of BIG-Bench that earlier models found hardest — multi-step reasoning, logic, and word problems. It's a long-standing standard for tracking reasoning ability across model generations.
58.5%
58.5%
median 50.8%open leader 87.5%
Overall best: Gemini 1.5 Pro 001 89.2%
70th percentile · #12 of 38 open · #17 of 50 all
#12 of 38
#17 of 50
Lab-reported
Epoch AI · Jul 2023

Llama 2 70B Chat HF math benchmarks

Symbolic and quantitative problem-solving, from competition math to multi-step calculation.

AIME 2024/2025
AIME (American Invitational Mathematics Examination) is a prestigious high-school competition of hard, integer-answer problems. It's a widely-cited yardstick for multi-step mathematical reasoning, here on the 2024–2025 papers.
0.0%
0.0%
median 52.8%open leader 98.6%
Overall best: Claude Fable 5 (high) 100.0%
lowest · #86 of 86 open · #283 of 283 all
#86 of 86
#283 of 283
Independent
Epoch AI · Jul 2023
MATH Level 5
MATH Level 5 covers the hardest tier of competition-mathematics problems, testing multi-step symbolic and quantitative reasoning.
3.3%
3.3%
median 41.6%open leader 96.6%
Overall best: GPT 5 (Aug 07, 2025, high) 98.1%
lowest · #39 of 39 open · #107 of 107 all
#39 of 39
#107 of 107
Independent
Epoch AI · Jul 2023

Other results (1)· leaderboard-only, not part of the llmrun Score

GSM8K
GSM8K is a classic set of grade-school math word problems requiring several arithmetic reasoning steps. Long a default sanity check for reasoning, top models now score near the ceiling, so it best separates smaller models.
58.7%
58.7%
median 52.3%open leader 94.5%
59th percentile · #26 of 62 open · #34 of 93 all
#26 of 62
#34 of 93
Lab-reported
Epoch AI · Jul 2023

Llama 2 70B Chat HF knowledge benchmarks

Breadth and depth of factual knowledge across many subjects, paired with the reasoning to apply it.

MMLU
MMLU (Massive Multitask Language Understanding) spans 57 subjects from history to law to medicine as multiple-choice questions. It is the long-standing default for broad knowledge and remains the most widely-reported general benchmark.
59.9%
59.9%
median 62.5%open leader 87.2%
Overall best: GPT 4o (Nov 20, 2024) 88.1%
45th percentile · #47 of 85 open · #90 of 136 all
#47 of 85
#90 of 136
Lab-reported
Epoch AI · Jul 2023

Llama 2 70B Chat HF instruction following benchmarks

How precisely a model obeys explicit, verifiable instructions — format, length, and content constraints — and, more broadly, how much people prefer its responses head-to-head in blind human voting.

LMArena Text
Arena (arena.ai, formerly LMArena / Chatbot Arena) ranks models by millions of blind, head-to-head human votes on real prompts. The Arena Score shown here is a Bradley-Terry rating computed with Arena's default style control, so higher means people preferred that model's answers more often — it's a relative rating, not a percentage, so compare it only to other models on this board.
1170.7
1170.7
median 1321.5open leader 1487.1
Overall best: Gemini 4 Argon (High) 1523.9
14th percentile · #161 of 187 open · #357 of 405 all
#161 of 187
#357 of 405
Arena
Arena · Oct 2026

Score vs VRAM in the ≤192 GB tier

Llama 2 70B Chat HF
Other models
Best Score for its VRAM
x: VRAM (GB, log scale) · y: llmrun Score

#78 of 78 that fit ≤192 GB

Each dot is a model at the best quant that fits ≤192 GB. The line joins the models with the best Score for their VRAM; a dot below it is beaten by a model that needs no more memory.

Full ≤192 GB ranking →

Llama 2 70B Chat HF vs its neighbours

Same boards for the nearest open models per llmrun Score. The highest score in each row is marked.

Board	Llama 2 70B Chat HFthis model	Qwen3.8 27Bbest in classQ4_K_M · 18 GB	Mistral Large Instruct 2407nearest by ScoreQ4_K_M · 86 GB	C4ai Command A 03 2025nearest by ScoreQ4_K_M · 79 GB	Mixtral 8x22B Instruct v0.1nearest by ScoreQ4_K_M · 88 GB
llmrun Score	50.4	122.1	81.1	79.8	71.6
GPQA Diamond	26.3%	—	49.0%	—	34.1%
DTBench	41.6%	88.0%	61.2%	61.3%	55.1%
ForecastBench	51.4	—	57.1	—	56.3
AIME 2024/2025	0.0%	—	8.5%	—	—
MATH Level 5	3.3%	—	44.8%	—	24.2%
MMLU	59.9%	—	80.0%	—	—
LMArena Text	1170.7	1437.8	1314.4	1353.7	1229.7
Compare head to head		Compare	Compare	Compare	Compare

Scores aggregated from public benchmark sources (Epoch AI: 8 boards, Arena: 1 board) — each linked from the benchmark pages. llmrun does not run these benchmarks. Percentiles and ranks are among open models in the llmrun catalogue; the median and leader marks refer to the same field.