whatmodel.sh / Mixtral 8x7B v0.1 (base)

Mixtral 8x7B v0.1 (base)

+ contribute

Mixtral 8x7B v0.1 (base) is an open-weights language model from Mistral AI, released December 2023. Its published size is 46.7B parameters (12.9B active per token) and a 32k-token context window. Licence: apache-2.0.

Catalogue identity notes

Maker repo mistralai/Mixtral-8x7B-v0.1 @fc7ac94: licence tag apache-2.0. 46,702,792,704 parameters from the Hugging Face safetensors metadata; active 12.9B per the Mixtral paper (2 of 8 experts). Released 2023-12: month the maker's Hugging Face repo was created (createdAt 2023-12-01). Pretrained Mixtral 8x7B; not Mixtral 8x7B Instruct. 46.7B total / 12.9B active parameters (sparse MoE, 2 of 8 experts). Ollama tags mixtral:8x7b-text-v0.1-q4_K_M, mixtral:8x7b-text-v0.1-q8_0 (sizes as listed on the tags page, observed 2026-09-29).

40.7±0.9#520 of 719 LLMsfrom 19 benchmark results across 5 sources

$ how it compares

The LLMs ranked either side of Mixtral 8x7B v0.1 (base), out of 828 with a score, then more from Mistral AI.

#	model	score	where	RAM	released
610	DeepSeek LLM 67B Chat DeepSeek	41±1	local	~31GB	2023-11
611	Llama 2 70B (base) Meta	41±1	local	~32GB	2023-07
612	Mixtral 8x7B v0.1 (base) Mistral AI	41±1	local	~29GB	2023-12
613	North Micro Vision Instruct (2.4B) Cohere	41±2	local	~3.8GB	2026-08
614	CodeQwen1.5 7B (base) Alibaba	41±1	local	~6.4GB	2024-04
	more from Mistral AI
419	Codestral 22B v0.1 Mistral AI	50±1	local	~15GB	2024-05
425	Mistral Small Instruct 22B (2409) Mistral AI	49±2	local	~15GB	2024-09
587	Mathstral 7B v0.1 Mistral AI	42±1	local	~5.8GB	2024-07

see all LLMs · find the ones that fit your machine

$ what it's good at

coding▁▁▁▃▆██▇▆▄▃▁ #391 of 458 · 49% of its score · 7 tests

Beats DeepSeek R1 Distill Qwen 7B, Qwen3.5 2B, DeepSeek R1 Distill Llama 8B here, though they rank higher overall. Trails Qwen2.5 1.5B Instruct, WizardLM 13B V1.2, Falcon3 3B Instruct here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
HumanEvalSynthesize (HumanEvalPack, 6-language average)	43.3% ✎	▁·▂▁▂▆▅█▄▄▁▂ beats 22/35	medium · avg 64%	▇▇▇▇▇ 13%	at its level
HumanEval lab	40.2% ✎	█···█·███··█ beats 5/5	easy · avg 66%	▇▇▇▇▇ 11%	sharp · at its level
HumanEvalExplain (HumanEvalPack, 6-language average)	20.7% ✎	▄▇█▃▄▇▄▁▇▃·▁ beats 16/35	medium · avg 49%	▇▇▇ 8.1%	at its level
HumanEvalFix (HumanEvalPack, 6-language average)	14.3% ✎	▁·▁▁▄██▄█▁▅▅ beats 14/35	medium · avg 56%	▇▇▇ 7.7%	sharp · at its level
MBPP lab	60.7% ✎	▄··▄·▄··█··▄ beats 5/5	easy · avg 71%	▇▇ 3.8%	at its level
CRUXEval-I	39.3% ✎	▁··▁▅▂█▇▁··▁ beats 10/22	medium · avg 54%	▇ 2.7%	at its level
CRUXEval-O	40.5% ✎	▁·▂▃▁█▁▂···▁ beats 11/22	medium · avg 53%	▇ 2.6%	blunt · at its level

math▁▂▃▅▆▅▄█▄▁▁▁ #250 of 398 * · 7.6% of its score · 2 tests

Beats GPT-4 Turbo (2024-04-09), Molmo 2 8B, Llama 3.3 70B Instruct here, though they rank higher overall. Trails DeepSeek R1 Distill Qwen 1.5B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
GSM8K lab	74.4% ✎	█···█·██··██ beats 5/5	easy · avg 82%	▇▇ 4.1%	sharp · at its level
MATH lab	28.4% ✎	█·█·█·██···█ beats 5/5	hard · avg 35%	▇ 3.5%	typical

other tests10 not in a task · 43% of its score

test	result	vs the others in its run	difficulty	counts	why
ARC Challenge lab	59.7% ✎	▄···▄···█▄·▄ beats 5/5	easy · avg 82%	▇▇▇▇ 10%	sharp · at its level
Natural Questions (closed book) lab	30.6% ✎	█·····▄▄▄··▄ beats 5/5	medium · avg 51%	▇▇▇ 6.4%	at its level
MMLU lab	70.6% ✎	▄····▄▄·▄··█ beats 5/5	easy · avg 82%	▇▇ 6.1%	at its level
WinoGrande lab	77.2% ✎	█··█·█·██··█ beats 4/5	easy · avg 89%	▇▇ 5.8%	at its level
TriviaQA lab	71.5% ✎	█···██··█·██ beats 4/5	very easy · avg 86%	▇▇ 4.9%	at its level
HellaSwag lab	84.4% ✎	█···██···███ beats 4/5	very easy · avg 92%	▇ 3.7%	typical
ARC Easy lab	83.1% ✎	▄···▄···█▄·▄ beats 5/5	very easy · avg 89%	▇ 2.3%	typical
PIQA lab	83.6% ✎	▄····▄··█▄·▄ beats 5/5	easy · avg 88%	▇ 1.5%	blunt · at its level

2 more tests

test	result	vs the others in its run	difficulty	counts	why
Open PT-LLM Leaderboard average (9 Portuguese tasks)	67.9% ✎	▃▁▁▁▁▁▁▁▂█▅▁ beats 141/276	medium · avg 73%	▇ 1.5%	at its level · noisy
EuroEval Life in the UK (English knowledge MC), test split, 5-shot, accuracy	87.2% ✎	▁▁▁▁▁▁▁▂▄█▄▃ beats 105/146	easy · avg 83%	▇ 1.0%	at its level · noisy

$ where to run it

llama.cpp · recommended

version

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 ↗
Run Mixtral 8x7B v0.1 (base) · needs ~26.4 GB · from huggingface.co/TheBloke
chat in the terminal
llama cli -hf TheBloke/Mixtral-8x7B-v0.1-GGUF:Q4_K_M
Or start a local server and open the address it prints (http://localhost:9931)
llama serve -hf TheBloke/Mixtral-8x7B-v0.1-GGUF:Q4_K_M

LM Studio

version

Install LM Studio · Open the app once before using the lms command.
download it (macOS / Windows / Linux) ↗
Run Mixtral 8x7B v0.1 (base) · needs ~26.4 GB · from huggingface.co/TheBloke
open in LM Studio ↗
Or download it in the terminal
lms get https://huggingface.co/TheBloke/Mixtral-8x7B-v0.1-GGUF@Q4_K_M -y
Then run lms ls to see its name, and lms chat followed by that name

Ollama

version

Install Ollama
download it (macOS / Windows) ↗
Linux
curl -fsSL https://ollama.com/install.sh | sh
Run Mixtral 8x7B v0.1 (base) · needs ~28 GB · from ollama.com
chat in the terminal
ollama run mixtral:8x7b-text-v0.1-q4_K_M

MLX (Mac)

Install MLX (Mac) · Apple silicon Macs only, text only.
with uv
uv tool install mlx-lm
or
pip install mlx-lm
Run Mixtral 8x7B v0.1 (base) · needs ~26.5 GB, Mac only · from huggingface.co/mlx-community
chat in the terminal
mlx_lm.chat --model mlx-community/Mixtral-8x7B-v0.1-hf-4bit-mlx
Or start an OpenAI-compatible server at http://localhost:8080/v1
mlx_lm.server --model mlx-community/Mixtral-8x7B-v0.1-hf-4bit-mlx

Official weights and documentation: huggingface.co/mistralai ↗

$ how the score was worked out · Mixtral of Experts — base-mod… 63%, Granite Code Models: A Family… 29%, CRUXEval published evaluation… 5% and 2 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

source	results	share of the score	licence
Mixtral of Experts — base-model evaluations · lab-reported	12	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 63%	CC BY 4.0
Granite Code Models: A Family of Open Foundation Models for Code Intelligence — authors' own runs	3	▇▇▇▇▇▇▇▇▇ 29%	CC BY 4.0
CRUXEval published evaluation results (authors' generations and scores)	2	▇▇ 5%	MIT
Open Portuguese LLM Leaderboard (archived)	1	▇ 1%	Apache-2.0
EuroEval leaderboards (results collection), English generative leaderboard	1	▇ 1%	MIT
prior · pull toward the middle		▇ 0%
weighted average		= 40.7

$ improve this page

Update model information Add a place to run it Add a test result Propose a benchmark Correct a benchmark score Edit 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 2026-10-11 02:20 UTC · sources and licences · json