[whatmodel.sh](../../README.md) / Llama 2 13B Chat
# Llama 2 13B Chat
Meta · released 2023-07 · 13.02B params · llama2
**score 35.5** ±0.8 · #622 of 732 · from 30 benchmark results across 16 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks Llama 2 13B Chat among every model on that task or test, worst to best.
task vs the others rank counts tests
coding ▁▁●▃▅▇█▇▅▄▃▁ #455 of 469* 4.8% 1
chat & writing ▁▁▁●▆▄▆█▆▇▇▃ #365 of 402 33% 6
math ▁●▃▅█▆▄█▄▂▁▁ #416 of 431* 10% 1
coding · #455 of 469 · 4.8% of its score
Beats [Molmo 7B-D (0924)](molmo-7b-d.md), [DeepSeek R1 Distill Qwen 1.5B](deepseek-r1-distill-qwen-1-5b.md), [Qwen3 0.6B](qwen3-0-6b.md) here, though they rank higher overall. Trails [Falcon3 1B Instruct](falcon3-1b.md), [Gemma 2B IT](gemma-2b.md), [Granite 3.0 1B-A400M Instruct](granite-3-0-1b-a400m.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
Tulu 2: CodexEval pass@10 33.1% ▃●▃·▆▃▃█▃▂▂▅ beats 3/25 easy · avg 83% 4.8% sharp · at its level
chat & writing · #365 of 402 · 33% of its score
Beats [Code Llama 70B Instruct](codellama-70b-instruct.md), [Gemma 4 E2B](gemma-4-e2b.md), [Code Llama 34B Instruct](codellama-34b-instruct.md) here, though they rank higher overall. Trails [Gemma 3 270M IT](gemma-3-270m.md), [SmolLM2 360M Instruct](smollm2-360m.md), [Granite 4.0 H 350M](granite-4-0-h-350m.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
AlpacaEval 2 length-controlled 8.4% ▁▂▂●▆▆█▆▄▄▁▁ beats 18/122 medium · avg 35% 11% sharp
LMArena (text) 1141.5 ▁▁▁●▃▃▃▆▆▇█▃ beats 33/385 avg 1260 7.1% blunt
Polyrating rating (bias-aware re-… 1081.0 ▂▁▂▁▃▇●▆▇▄▂▂ beats 38/95 avg 1187 6.5% typical
MT-Bench (GPT-4 single-answer gra… 6.6500 MT-Bench score (1-10) ▆▂▂▆▆▄█●▂▄▄▆ beats 19/28 avg 8.539 MT-Bench scor… 5.0% sharp · noisy
Arabic IFEval prompt-level (Incep… 2.0% ▁·●▁▂▁▄▄▆▄█▄ beats 1/38 medium · avg 46% 2.5% sharp
AlpacaEval 1.0 win rate (vs text-… 81.1% ▂▂·▆·▇●▆█▆▄▂ beats 19/45 very easy · avg 94% 1.7% sharp · at its level · noisy
math · #416 of 431 · 10% of its score
Trails [GPT-Neo 125M](gpt-neo-125m.md), [OPT-125M](opt-125m.md), [GPT-2 XL (1.5B)](gpt-2-xl.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
GSM8K 9.0% ●▂▅█▅▅▃▆···▂ beats 0/25 easy · avg 80% 10% sharp · at its level · several runs
other tests (19)
test result vs the others in its run difficulty counts why
BIG-Bench Hard 40.3% ▁·····●█▂▅·▁ beats 3/25 easy · avg 74% 13% sharp · several runs
MMLU 53.2% ▁▁▃▃▄●▄▁█··▁ beats 13/25 medium · avg 73% 7.1% at its level · several runs
Natural Questions (closed book) 27.3% ▁·▁·▁▂▁▂·▅▂● beats 24/28 medium · avg 48% 6.2% typical
ARC Challenge 61.2% ▄▁▁▂▁▄█▇●▂▃▂ beats 108/140 easy · avg 82% 4.1% sharp · at its level
Open Persian LLM Leaderboard v2 a… 29.4% ●▄▂▃█▄▃▆▃▅▂▂ beats 0/81 medium · avg 57% 3.1% blunt · at its level
VEC embodied concepts, zero-shot … 55.0% ▁····▁▁·▁·●█ beats 5/11 hard · avg 63% 3.0% typical
AGIEval English 38.9% ●▄▄··▄··▄█·█ beats 1/9 easy · avg 65% 2.9% at its level
BRIDGE clinical text, zero-shot a… 20.9% ▁▁▁▁▁●▂▃▄▆█▃ beats 13/97 hard · avg 31% 2.8% blunt
EQ-Bench v1 (60 questions, paper … 33.0% ▂··▄·●▆▄▂█·▂ beats 3/14 medium · avg 53% 2.7% at its level
Open PT-LLM Leaderboard average (… 58.0% ▃▁▁▁▁▁▁▁●█▅▁ beats 84/276 medium · avg 72% 1.4% at its level · noisy
EuroEval Life in the UK (English … 57.6% ▁▁▁▁▁▁▁●▄█▄▃ beats 30/146 easy · avg 82% 1.2% at its level · noisy
ARC Easy 76.3% ▁▃▂▄▆█●▃▁··▁ beats 82/100 easy · avg 85% 1.1% blunt · at its level
Tulu 2: TyDiQA gold passage, 1-sh… 32.1% ▂··●▃▂▂▂█▅▁▂ beats 2/25 medium · avg 59% 1.0% blunt · at its level
Vectara hallucination leaderboard… 89.5% ▁▁▂●▄▄█▆▃▂▂▁ beats 23/140 very easy · avg 94% 0.8% blunt · noisy
TriviaQA 9.8% ▁··▁●▃▃▁▄▅▄█ beats 7/28 easy · avg 73% 0.8% at its level · noisy
HaluEval QA hallucination detecti… 57.3% █▁▁·▁▁·▁●▁▁▁ beats 22/27 hard · avg 66% 0.5% noisy
TruthfulQA MC2 44.0% ▅▃█●▁▂▅···▁▁ beats 16/31 medium · avg 57% 0.2% blunt · at its level
VEC visual concepts, zero-shot pe… 64.9% ▅▅▃█·●··▃·▃▃ beats 8/11 hard · avg 66% 0.0% blunt · at its level
Chess Puzzles (Epoch) 0.0% ●▁▁▁▁▂▃▃▃▂▁▁ beats 9/137 very hard · avg 5.8% 0.0% sharp · too hard for it · noisy
source results share of the score
Orca 2: Teaching Small Language Models … 6 22% ▇▇▇
Tulu 2 5 21% ▇▇▇
AlpacaEval 1.0 and 2.0 leaderboards 2 12% ▇
Hallucinations Leaderboard results 4 8% ▇
LMArena leaderboard dataset 1 7% ▇
Polyrating: A Cost-Effective and Bias-A… 1 7% ▇
LMSYS Chatbot Arena leaderboard Space: … 1 5% ▇
Open Persian LLM Leaderboard v2 1 3% ▇
Can Language Models Understand Physical… 2 3% ▇
BRIDGE Medical Leaderboard 1 3% ▇
EQ-Bench: An Emotional Intelligence Ben… 1 3% ▇
Inception Arabic/Hindi leaderboards 1 2% ▇
Open Portuguese LLM Leaderboard 1 1% ▇
EuroEval leaderboards 1 1% ▇
Vectara hallucination leaderboard 1 1% ▇
Epoch AI 1 0% ▇
prior: pull toward the middle 0% ▇
weighted average = 35.5
Licences: Orca 2: Teaching Small Language Models How to Reason ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Tulu 2 ([CC BY 4.0](../../licenses/CC-BY-4.0.txt)) · AlpacaEval 1.0 and 2.0 leaderboards ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)) · Hallucinations Leaderboard results ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Polyrating: A Cost-Effective and Bias-Aware Rating System for LLM Evaluation ([CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)) · LMSYS Chatbot Arena leaderboard Space: MT-Bench column of the published leaderboard table ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)) · Open Persian LLM Leaderboard v2 ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · Can Language Models Understand Physical Concepts? ([CC BY 4.0](../../licenses/acl-2023-emnlp-main-726-CC-BY-4-0.txt)) · BRIDGE Medical Leaderboard ([MIT](../../licenses/bridge-medical-leaderboard-MIT.txt)) · EQ-Bench: An Emotional Intelligence Benchmark for Large Language Models ([CC BY 4.0](../../licenses/paper-2312.06281v2-CC-BY-4.0.txt)) · Inception Arabic/Hindi leaderboards ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · Open Portuguese LLM Leaderboard ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · EuroEval leaderboards ([MIT](https://github.com/EuroEval/leaderboards/blob/1a77f1d2f8ba1ba05ab73287fb090d76bf41b2e0/LICENSE)) · Vectara hallucination leaderboard ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)) · Epoch AI ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/))
## How to run it
### llama.cpp (recommended)
1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md)
```sh
# macOS / Linux
curl -LsSf https://llama.app/install.sh | sh
# Windows
irm https://llama.app/install.ps1 | iex
```
2. Run Llama 2 13B Chat (Q4_K_M, needs ~7.9 GB, from [huggingface.co/TheBloke](https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF/tree/4458acc949de0a9914c3eab623904d4fe999050a))
```sh
llama cli -hf TheBloke/Llama-2-13B-chat-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf TheBloke/Llama-2-13B-chat-GGUF:Q4_K_M
```
Other versions: Q8_0 (~13.8 GB) `llama cli -hf TheBloke/Llama-2-13B-chat-GGUF:Q8_0`
### LM Studio
1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download)
2. Run Llama 2 13B Chat (Q4_K_M, needs ~7.9 GB, from [huggingface.co/TheBloke](https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF/tree/4458acc949de0a9914c3eab623904d4fe999050a))
```sh
lms get https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: Q8_0 (~13.8 GB) `lms get https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF@Q8_0 -y`
### Ollama
1. Install Ollama. [Download it (macOS / Windows)](https://ollama.com/download)
```sh
# Linux
curl -fsSL https://ollama.com/install.sh | sh
```
2. Run Llama 2 13B Chat (Q4_0, needs ~8.31 GB, from [ollama.com](https://ollama.com/library/llama2/tags))
```sh
ollama run llama2:13b-chat
```
Other versions: Q8_0 (~16.12 GB) `ollama run llama2:13b-chat-q8_0`
Model repository and documentation: [huggingface.co/meta-llama](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf)
[interactive version](https://whatmodel.sh/model/llama-2-13b-chat) · [JSON](https://whatmodel.sh/api?format=json&all=1)