[whatmodel.sh](../../README.md) / Llama 2 70B Chat
# Llama 2 70B Chat
Meta · released 2023-07 · 68.98B params · llama2
**score 40.6** ±0.7 · #554 of 732 · from 31 benchmark results across 19 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 70B Chat among every model on that task or test, worst to best.
task vs the others rank counts tests
coding ▁▁▁●▅▇█▇▅▄▃▁ #430 of 469* 3.3% 1
chat & writing ▁▁▁▃●▄▆█▆▇▇▃ #342 of 402 36% 7
math ▁▁●▅█▆▄█▄▂▁▁ #399 of 431 14% 4
coding · #430 of 469 · 3.3% of its score
Beats [DeepSeek R1 Distill Qwen 7B](deepseek-r1-distill-qwen-7b.md), [Qwen3.5 2B](qwen3-5-2b.md), [DeepSeek R1 Distill Llama 8B](deepseek-r1-distill-llama-8b.md) here, though they rank higher overall. Trails [Llama 3.2 1B Instruct](llama-3-2-1b.md), [SmolLM2 1.7B Instruct](smollm2-1-7b.md), [Mistral 7B Instruct v0.1](mistral-7b-instruct-v0-1.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
Tulu 2: CodexEval pass@10 52.1% ▃▃▃·●▃▃█▃▂▂▅ beats 8/25 easy · avg 83% 3.3% sharp · at its level
chat & writing · #342 of 402 · 36% of its score
Beats [OpenThinker3 7B](openthinker3-7b.md), [Code Llama 70B Instruct](codellama-70b-instruct.md), [Gemma 4 E2B](gemma-4-e2b.md) here, though they rank higher overall. Trails [Granite 4.0 H 350M](granite-4-0-h-350m.md), [LFM2 350M](lfm2-350m.md), [Gemma 3 1B](gemma-3-1b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
AlpacaEval 2 length-controlled 14.7% ▁▂▂▃●▆█▆▄▄▁▁ beats 36/122 medium · avg 35% 12% sharp
Arena-Hard (v0.1, GPT-4-1106 judg… 11.6% ▂▁●▅█▅▄▃▃▅▃▁ beats 5/60 medium · avg 47% 6.2% sharp
LMArena (text) 1170.7 ▁▁▁▂●▃▃▆▆▇█▃ beats 49/385 avg 1260 5.2% blunt
Polyrating rating (bias-aware re-… 1108.0 ▂▁▂▁▃▇█●▇▄▂▂ beats 51/95 avg 1187 4.8% typical
MT-Bench (GPT-4 single-answer gra… 6.8600 MT-Bench score (1-10) ▆▂▂▆▆▄█●▂▄▄▆ beats 20/28 avg 8.539 MT-Bench scor… 3.7% sharp · noisy
Arabic IFEval prompt-level (Incep… 3.0% ▁·▁●▂▁▄▄▆▄█▄ beats 2/38 medium · avg 46% 2.8% sharp
AlpacaEval 1.0 win rate (vs text-… 92.7% ▂▂·▆·▇▃▆█●▄▂ beats 35/45 very easy · avg 94% 0.9% sharp · noisy
math · #399 of 431 · 14% of its score
Beats [Command R (03-2024)](command-r-03-2024.md), [Molmo 7B-D (0924)](molmo-7b-d.md), [Mistral 7B Instruct v0.3](mistral-7b-instruct-v0-3.md) here, though they rank higher overall. 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 59.0% ▆▂▅█▅▅●▆···▂ beats 19/25 easy · avg 80% 8.2% sharp · at its level · several runs
MGSM 34.1% ▂·▁▁▁▁▁▂●█▃▂ beats 14/42 medium · avg 63% 3.4% sharp · at its level · small run
MATH level 5 3.3% ●▂▅▆▄█▆▃▅▂▃▄ beats 0/92 hard · avg 28% 2.2% sharp · too hard for it
OTIS mock AIME 2024–2025 0.0% ●▃▄▃▃▃▇▇█▃▂▅ beats 1/197 very hard · avg 7.2% 0.1% sharp · too hard for it · noisy
other tests (16)
test result vs the others in its run difficulty counts why
BIG-Bench Hard 49.0% ▁·····▂●▂▅·▁ beats 10/25 easy · avg 74% 12% sharp · at its level · several runs
MMLU 60.9% ▁▁▃▃▄█▄●█··▁ beats 18/25 medium · avg 73% 6.2% at its level · several runs
WildBench v2 WB-Score (task macro) 20.6596 WB-Score ▁·▁▁·●█▄▇▆▅▆ beats 3/52 avg 43.019 WB-Score 5.5% sharp
F2WTQ free-form table QA, 2-shot … 37.6% ▄···█·▄●··▄▄ beats 5/7 medium · avg 56% 4.6% at its level
ARC Challenge 67.7% ▄▁▁▂▁▄█▇●▂▃▂ beats 114/140 easy · avg 82% 3.2% sharp · at its level
Open Persian LLM Leaderboard v2 a… 32.6% ●▄▂▃█▄▃▆▃▅▂▂ beats 3/81 medium · avg 57% 2.4% blunt · at its level
BRIDGE clinical text, zero-shot a… 25.1% ▁▁▁▁▁▂●▃▄▆█▃ beats 18/97 hard · avg 31% 2.4% blunt
AGIEval English 46.7% █●▄··▄··▄█·█ beats 2/9 easy · avg 65% 2.2% at its level
EQ-Bench v1 (60 questions, paper … 51.6% ▂··▄·▂▆▄▂●·▂ beats 12/14 medium · avg 53% 2.2% at its level
GPQA Diamond 26.3% ▁●▁▁▂▄▆▄▅▇█▅ beats 7/211 hard · avg 39% 1.8% sharp · too hard for it
Open PT-LLM Leaderboard average (… 65.7% ▃▁▁▁▁▁▁▁▂●▅▁ beats 118/276 medium · avg 72% 1.2% at its level · noisy
EuroEval Life in the UK (English … 57.5% ▁▁▁▁▁▁▁●▄█▄▃ beats 29/146 easy · avg 82% 0.9% at its level · noisy
Tulu 2: TyDiQA gold passage, 1-sh… 44.4% ▂··▁▃▂●▂█▅▁▂ beats 8/25 medium · avg 59% 0.8% blunt · at its level
ARC Easy 82.2% ▁▃▂▄▆█▅●▁··▁ beats 94/100 easy · avg 85% 0.7% blunt · at its level
Vectara hallucination leaderboard… 94.1% ▁▁▂▃▄●█▆▃▂▂▁ beats 51/140 very easy · avg 94% 0.5% blunt · noisy
FANToM BeliefQ [Choice], short co… 38.4% ●·····▂····▁ beats 3/11 very hard · avg 55% 0.2% sharp · too hard for it · noisy
source results share of the score
Orca 2: Teaching Small Language Models … 6 19% ▇▇
Tulu 2 5 18% ▇▇
AlpacaEval 1.0 and 2.0 leaderboards 2 13% ▇▇
Arena-Hard-Auto v0.1 leaderboard 1 6% ▇
WildBench v2 leaderboard data 1 6% ▇
LMArena leaderboard dataset 1 5% ▇
Polyrating: A Cost-Effective and Bias-A… 1 5% ▇
Investigating Table-to-Text Generation … 1 5% ▇
Epoch AI 3 4% ▇
LMSYS Chatbot Arena leaderboard Space: … 1 4% ▇
Not All Languages Are Created Equal in … 1 3% ▇
Inception Arabic/Hindi leaderboards 1 3% ▇
Open Persian LLM Leaderboard v2 1 2% ▇
BRIDGE Medical Leaderboard 1 2% ▇
EQ-Bench: An Emotional Intelligence Ben… 1 2% ▇
Open Portuguese LLM Leaderboard 1 1% ▇
EuroEval leaderboards 1 1% ▇
Vectara hallucination leaderboard 1 1% ▇
FANToM: A Benchmark for Stress-testing … 1 0% ▇
prior: pull toward the middle 0% ▇
weighted average = 40.6
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)) · Arena-Hard-Auto v0.1 leaderboard ([Apache-2.0](../../licenses/arena-hard-auto-v0-1-Apache-2.0.txt)) · WildBench v2 leaderboard data ([Apache-2.0](https://github.com/allenai/WildBench/blob/d6b8dcaf377d173d031980f97c16e1a82618c03d/LICENSE)) · 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/)) · Investigating Table-to-Text Generation Capabilities of Large Language Models in Real-World Information Seeking Scenarios ([CC BY 4.0](../../licenses/acl-2023-emnlp-industry-17-CC-BY-4-0.txt)) · Epoch AI ([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)) · Not All Languages Are Created Equal in LLMs: Improving Multilingual Capability by Cross-Lingual-Thought Prompting ([CC BY 4.0](../../licenses/acl-2023-findings-emnlp-826-CC-BY-4-0.txt)) · Inception Arabic/Hindi leaderboards ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · Open Persian LLM Leaderboard v2 ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.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)) · 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)) · FANToM: A Benchmark for Stress-testing Machine Theory of Mind in Interactions ([CC BY 4.0](../../licenses/acl-2023-emnlp-main-890-CC-BY-4-0.txt))
## 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 70B Chat (Q4_K_M, needs ~41.4 GB, from [huggingface.co/TheBloke](https://huggingface.co/TheBloke/Llama-2-70B-Chat-GGUF/tree/798ec30fbdbdde983a1189c8396ea41b8980aa16))
```sh
llama cli -hf TheBloke/Llama-2-70B-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-2-70B-Chat-GGUF:Q4_K_M
```
Other versions: Q2_K (~29.3 GB) `llama cli -hf TheBloke/Llama-2-70B-Chat-GGUF:Q2_K`
### 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 70B Chat (Q4_K_M, needs ~41.4 GB, from [huggingface.co/TheBloke](https://huggingface.co/TheBloke/Llama-2-70B-Chat-GGUF/tree/798ec30fbdbdde983a1189c8396ea41b8980aa16))
```sh
lms get https://huggingface.co/TheBloke/Llama-2-70B-Chat-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: Q2_K (~29.3 GB) `lms get https://huggingface.co/TheBloke/Llama-2-70B-Chat-GGUF@Q2_K -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 70B Chat (Q4_0, needs ~41.89 GB, from [ollama.com](https://ollama.com/library/llama2/tags))
```sh
ollama run llama2:70b-chat
```
Other versions: Q8_0 (~83.28 GB) `ollama run llama2:70b-chat-q8_0`
Model repository and documentation: [huggingface.co/meta-llama](https://huggingface.co/meta-llama/Llama-2-70b-chat-hf)
[interactive version](https://whatmodel.sh/model/llama-2-70b-chat) · [JSON](https://whatmodel.sh/api?format=json&all=1)