whatmodel.sh / Hermes 3 Llama 3.1 70B

Hermes 3 Llama 3.1 70B

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Hermes 3 Llama 3.1 70B is an open-weights language model from Nous Research, released July 2024. Its published size is 70.55B parameters and a 131k-token context window. Licence: llama3.

Catalogue identity notes

NousResearch/Hermes-3-Llama-3.1-70B @1fc86da (created 2024-07-29, licence llama3). Nous Research Hermes 3 70B (Aug 2024), full-parameter fine-tune of Llama 3.1 70B; sibling of the catalogued hermes-3-llama-3-1-8b. Lab rows from the Hermes 3 report Table 5 (rule b; 405B evaluated under FP8); independent LiveBench 2024-07-26 row.

Model information sources

Release date ↗ · Weights licence ↗

HF safetensors metadata 70,553,706,496 parameters (NousResearch/Hermes-3-Llama-3.1-70B @1fc86da).

48.5±1.0#374 of 719 LLMsfrom 14 benchmark results across 2 sources

$ how it compares

The LLMs ranked either side of Hermes 3 Llama 3.1 70B, out of 828 with a score, then more from Nous Research.

#modelscorewhereRAMreleased
445Magicoder-S-DS 6.7B UIUC (ise-uiuc)49±1local~5.8GB2023-12
446InternLM2.5 7B Chat Shanghai AI Laboratory (InternLM)49±2local~6.1GB2024-06
447Hermes 3 Llama 3.1 70B Nous Research49±1local~44GB2024-07
448Qwen2.5 7B Instruct Alibaba48±1api $0.30/M +local~6GB2024-09
449DeepSeek R1 Distill Qwen 7B DeepSeek48±1local~6GB2025-01
more from Nous Research
389Hermes 3 Llama 3.1 405B Nous Research51±1local~114GB2024-08
461Hermes 2 Theta Llama-3 70B Nous Research48±2local~47GB2024-06
498Hermes 2 Pro Llama 3 8B Nous Research46±3local~6.6GB2024-04

see all LLMs · find the ones that fit your machine

$ what it's good at

chat & writing█········█·█ #2 of 3 * · 1.8% of its score · 1 test
testresultvs the others in its rundifficultycountswhy
IFEval (aggregation unspecified) lab81.2% ✎▁▁▃▁▂▄▄▆█▃▂▁ beats 55/141easy · avg 79%▇ 1.8%small run
other tests13 not in a task · 98% of its score
testresultvs the others in its rundifficultycountswhy
LiveBench 2024-07-2640.1% ✎▁▁▁▂▂▃▂█▆▃▅▃ beats 44/61medium · avg 41%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 40%at its level
BIG-Bench Hard lab67.8% ✎▁▁▁▁▂▄▅▅█▃▄▁ beats 84/129medium · avg 72%▇▇▇▇ 9.3%sharp · at its level · small run
GPQA (subset unspecified) lab37.7% ✎▂▁▂▂▄▃▆▃▅█▄▃ beats 33/104hard · avg 41%▇▇▇ 8.2%sharp · small run
ARC Challenge lab65.5% ✎▆▁▁▂▂▅▇█▄▃▃▂ beats 77/97medium · avg 73%▇▇▇ 7.7%sharp · at its level · small run
MMLU lab79.1% ✎▁▁▁▁▁▂▃▅█▃▆▃ beats 133/183easy · avg 80%▇▇ 5.8%at its level · small run
WinoGrande lab83.2% ✎▁···▂▁▁▅█▇▃▃ beats 84/87easy · avg 86%▇▇ 5.1%at its level · small run
MMLU-Pro lab47.2% ✎▁▁·▁▁▁▃▂▄▅█▁ beats 42/188medium · avg 52%▇▇ 5.1%at its level · small run
OpenBookQA lab49.4% ✎█▁▁▁▄▃▃▇▂▁▄▂ beats 46/65medium · avg 55%▇▇ 4.3%at its level · small run
5 more tests
testresultvs the others in its rundifficultycountswhy
BoolQ lab88.0% ✎▁▄▅▆▂▆▄▄█▃▄▄ beats 55/58easy · avg 90%▇▇ 4.1%small run
HellaSwag lab88.2% ✎▁▁▁▁▂▂▃▇█▆▅▃ beats 82/85very easy · avg 89%▇ 3.7%small run
MuSR lab50.7% ✎█······█···█ beats 2/2medium · avg 53%▇ 1.8%blunt · at its level · small run
PIQA lab84.4% ✎▁▁▁▃▃▄▇█▄▃▃▁ beats 80/83easy · avg 86%▇ 1.5%blunt · at its level · small run
TruthfulQA MC2 lab63.3% ✎█▅▇▅▁▄▁▇▅▄▁▁ beats 51/73medium · avg 64%▇ 0.8%blunt · at its level · small run

$ where to run it

llama.cpp · recommended

  1. Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.
    macOS / Linux
    curl -LsSf https://llama.app/install.sh | sh
    Windows
    irm https://llama.app/install.ps1 | iex

    other ways to install ↗

  2. Run Hermes 3 Llama 3.1 70B · needs ~42.52 GB · from huggingface.co/NousResearch (official)
    chat in the terminal
    llama cli -hf NousResearch/Hermes-3-Llama-3.1-70B-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf NousResearch/Hermes-3-Llama-3.1-70B-GGUF:Q4_K_M

LM Studio

  1. Install LM Studio · Open the app once before using the lms command.

    download it (macOS / Windows / Linux) ↗

  2. Run Hermes 3 Llama 3.1 70B · needs ~42.52 GB · from huggingface.co/NousResearch (official)

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B-GGUF@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run Hermes 3 Llama 3.1 70B · needs ~40 GB · from ollama.com
    chat in the terminal
    ollama run hermes3:70b

Official weights and documentation: huggingface.co/NousResearch ↗

$ how the score was worked out · Nous Research vendor-reported… 59%, LiveBench model judgments (HF… 40%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
Nous Research vendor-reported scores for its own Hermes models (Hermes 3 Technical Report) · lab-reported13▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 59%Factual vendor-reported score (not licensed content)
LiveBench model judgments (HF livebench/model_judgment, 2024-10-22 revision) with the LiveBench question datasets1▇▇▇▇▇▇▇▇▇▇▇▇ 40%Apache-2.0 (stated in the LiveBench datasheet)
prior · pull toward the middle▇ 0%
weighted average= 48.5

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json