Model

Llama 2 13B

OPEN

Meta · Llama 2 family · released Jul 18, 2023

Mid-size Llama 2 base model, pretrained on 2T tokens with a 4K context window and the first commercially-licensed Llama generation.

ReasoningCodingVisionFunction callingTool useAgentic

69.7

Elo · rank #394

Parameters

13B

Active params

13B (dense)

Context

4K tokens

Architecture

Dense decoder-only Transformer

License

Llama 2 Community License Agreement

Languages

—

API price (in/out)

No hosted API

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

AGIEvalReasoning39.1%#32

best: OLMo 3-Think 32B · 88.2%

ARC-ChallengeReasoning49.4%#127

best: Llama 3.1 405B · 96.9%

ARC-EasyReasoning77.3%#31

best: Phi-3-medium (14B) · 97.7%

BIG-Bench HardReasoning39.4%#114

best: ERNIE 4.5 300B-A47B · 94.3%

GSM8KMath28.7%#154

best: Llama 3.1 405B · 96.8%

HellaSwagReasoning80.7%#75

best: Claude 3 Opus · 95.4%

HumanEvalCoding18.3%#169

best: Claude Opus 4.5 · 99.4%

MATH-500Math3.9%#200

best: GPT-5 · 99.4%

MBPPCoding30.6%#96

best: Llama-3.3-Nemotron-Super-49B v1 (Reasoning On) · 91.3%

MMLUKnowledge54.8%#222

best: OpenAI o3 · 92.9%

OpenBookQAReasoning57.0%#18

best: Claude 1 · 90.8%

PIQAReasoning80.5%#37

best: GPT-4o mini · 93.1%

Social IQaReasoning50.3%#27

best: Apple DCLM-Baseline 7B · 82.9%

TriviaQAKnowledge73.1%#20

best: Sarvam-1 (2B) · 90.6%

TruthfulQAKnowledge41.9%#73

best: Phi-3.5-MoE (16x3.8B, 6.6B active) · 77.5%

WinoGrandeReasoning72.8%#81

best: PaLM 2 · 90.9%

Run it locally

VRAM @ Q4

10 GB

VRAM @ FP16

26 GB

Fits on (Q4)

RTX 3060 12GBRTX 4070 Ti 16GBRTX 3090 24GBRTX 4090 24GBRTX 5090 32GBM4 Pro 48GBM3 Max 128GBM3 Ultra 512GBA100 80GBH100 80GBH200 141GBB200 192GB

No verified base-model-specific RTX 4090 figure; the widely-cited ~71 tok/s Ollama benchmark for 'llama2:13b' targets Ollama's default chat-tuned weights, not this base checkpoint.

Quantizations

GGUF · GPTQ · AWQ · GGML

Llama 2 family

Elo progression across releases

Llama 2 13BJul 202369.7 Llama 2 13B ChatJul 2023105.1 Llama 2 70BJul 2023365.0 Llama 2 70B ChatJul 2023363.8 Llama 2 7BJul 2023-121.8 Llama 2 7B ChatJul 2023-10.6

Resources

HFmeta-llama/Llama-2-13b GHmeta-llama/llama DOCSai.meta.com/llama APIAvailable via Meta API

API price weights · each benchmark row carries its own source badge (see methodology)