whatmodel.sh / Snowflake Arctic Instruct

Snowflake Arctic Instruct

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

Snowflake Arctic Instruct is an open-weights language model from Snowflake, released April 2024. Its published size is 480B parameters (17B active per token). Licence: apache-2.0.

Catalogue identity notes

Comparison-only. Arena entry 'snowflake-arctic-instruct' (text). LMArena's leaderboard table (arena-leaderboard space @6f2eeeb) labels 'snowflake-arctic-instruct' as 'Snowflake Arctic Instruct' (Snowflake, Apache 2.0), linking https://www.snowflake.com/blog/arctic-open-efficient-foundation-language-models-snowflake/. Maker repo Snowflake/snowflake-arctic-instruct @be318ca: licence tag apache-2.0. Parameters: card: 480B total and 17B active parameters (top-2 gating MoE). Released 2024-04: month the maker's Hugging Face repo was created (createdAt 2024-04-21).

41.4±1.7#509 of 719 LLMsfrom 5 benchmark results across 4 sources

$ how it compares

The LLMs ranked either side of Snowflake Arctic Instruct, out of 828 with a score.

#modelscorewhereRAMreleased
599OLMo 2 7B Instruct (1124) Allen Institute for AI42±1local~5.8GB2024-12
600WizardCoder-15B-V1.0 WizardLM42±1local~12GB2023-06
601Snowflake Arctic Instruct Snowflake41±2local~1007GB2024-04
602Phi-3-vision-128k-instruct Microsoft41±2local~10GB2024-05
603OpenChat 3.5 OpenChat41±2local~6.1GB2023-10

see all LLMs · find the ones that fit your machine

$ what it's good at

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #319 of 393 · 97% of its score · 3 tests

Beats Gemma 4 E4B, OpenThinker3 7B, Gemma 4 E2B here, though they rank higher overall. Trails LFM2 350M, Gemma 3 1B, Flan-T5 XXL here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena-Hard (v0.1, GPT-4-1106 judge)17.6% ✎▂▁▁▅█▅▄▃▃▅▃▁ beats 11/60medium · avg 52%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 45%sharp
Polyrating rating (bias-aware re-fit of Chatbot Arena votes, mid-2024)1126.0 ✎▂▁▂▁▃▇█▆▇▄▂▂ beats 61/95avg 1201▇▇▇▇▇▇▇▇▇▇ 26%blunt
LMArena (text)1180.0 ✎▁▁▁▂▃▃▃▆▆▇█▃ beats 56/379avg 1268▇▇▇▇▇▇▇▇▇▇ 25%blunt
other tests2 not in a task · 2.2% of its score
testresultvs the others in its rundifficultycountswhy
Vectara hallucination leaderboard, 2025 dataset (HHEM-2.3, factual consistency rate)97.0% ✎▁▁▂▃▄▄█▆▃▂▂▁ beats 103/140very easy · avg 94%▇ 2.2%blunt · noisy
Vectara hallucination leaderboard (HHEM-2.3, factual consistency rate)95.7% ✎▁▂▁▂▅█▂▃▄▃▃▁ beats 95/98very easy · avg 90%▇ 0.0%blunt

$ where to run it

Other ways to run it

  1. Run it as published · needs ~957.17 GB, NVIDIA only · from huggingface.co/Snowflake (official)
    pip install 'transformers>=4.39.0' 'deepspeed>=0.14.2' accelerate && python -c "import torch; from transformers import AutoModelForCausalLM, AutoTokenizer; from deepspeed.linear.config import QuantizationConfig; r='Snowflake/snowflake-arctic-instruct'; t=AutoTokenizer.from_pretrained(r, trust_remote_code=True); m=AutoModelForCausalLM.from_pretrained(r, trust_remote_code=True, low_cpu_mem_usage=True, device_map='auto', ds_quantization_config=QuantizationConfig(q_bits=8), max_memory={i: '150GiB' for i in range(8)}, torch_dtype=torch.bfloat16); x=t.apply_chat_template([{'role': 'user', 'content': '5x + 35 = 7x - 60 + 10. Solve for x'}], add_generation_prompt=True, return_tensors='pt').to('cuda'); print(t.decode(m.generate(input_ids=x, max_new_tokens=256)[0]))"

Official weights and documentation: huggingface.co/Snowflake ↗

$ how the score was worked out · Arena-Hard-Auto v0.1 leaderbo… 45%, Polyrating: A Cost-Effective… 26%, LMArena leaderboard dataset 25% and 1 more

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
Arena-Hard-Auto v0.1 leaderboard (repository README, pre-v2)1▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 45%Apache-2.0
Polyrating: A Cost-Effective and Bias-Aware Rating System for LLM Evaluation (arXiv:2409.00696v3)1▇▇▇▇▇▇▇▇ 26%CC-BY-4.0
LMArena leaderboard dataset1▇▇▇▇▇▇▇▇ 25%CC BY 4.0
Vectara hallucination leaderboard (vectara/hallucination-leaderboard repository)2▇ 2%Apache-2.0
prior · pull toward the middle▇ 1%
weighted average= 41.4

$ 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