[whatmodel.sh](../../README.md) / Zephyr 7B beta # Zephyr 7B beta Hugging Face · released 2023-10 · 7.24B params · mit **score 35.4** ±0.6 · #623 of 732 · from 20 benchmark results across 9 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks Zephyr 7B beta among every model on that task or test, worst to best. 
task            vs the others          rank  counts  tests
coding          ▁▁▁●▅▇█▇▅▄▃▁    #436 of 469     59%      4
chat & writing  ▁▁▁●▆▄▆█▆▇▇▃    #369 of 402     15%      4
math            ▁▁▃▅●▆▄█▄▂▁▁   #331 of 431*    2.5%      1
coding · #436 of 469 · 59% 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
HumanEval                   30.0%  ▁·▁▁●▄▆▅█▃▁▁ beats 15/99  easy · avg 76%     31%  sharp · at its level · several runs
HumanEval+                  23.2%  ▁·▁▁▁●▄▄▄█▃▁ beats 11/99  easy · avg 70%     21%  sharp · at its level
MBPP+                       34.7%  ▁···▁●▂▅▇█▆▂ beats 3/75   easy · avg 66%    4.7%  at its level
Tulu 2: CodexEval pass@10   54.3%  ▃▃▃·●▃▃█▃▂▂▅ beats 9/25   easy · avg 83%    2.8%  sharp · at its level
chat & writing · #369 of 402 · 15% of its score Beats [Code Llama 70B Instruct](codellama-70b-instruct.md), [Gemma 4 E2B](gemma-4-e2b.md), [Phi-3-mini-128k-instruct (Apr 2024)](phi-3-mini-128k-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 350M](granite-4-0-350m.md) here, though they rank lower overall. 
test                                result  vs the others in its run   difficulty           counts  why
AlpacaEval 2 length-controlled       13.2%  ▁▂▂▃●▆█▆▄▄▁▁ beats 32/122  medium · avg 35%       6.1%  sharp
LMArena (text)                      1130.9  ▁▁▁●▃▃▃▆▆▇█▃ beats 28/385  avg 1260               4.1%  blunt
Polyrating rating (bias-aware re-…  1049.0  ▂▁▂▁▃●█▆▇▄▂▂ beats 25/95   avg 1187               3.8%  typical
AlpacaEval 1.0 win rate (vs text-…   90.6%  ▂▂·▆·▇▃▆●▆▄▂ beats 32/45   very easy · avg 94%    1.0%  sharp · at its level · noisy
math · #331 of 431 · 2.5% of its score Beats [GPT-4 (0613)](gpt-4-0613.md), [Athene-70B](athene-70b.md), [Llama 3 70B Instruct](llama-3-70b.md) here, though they rank higher overall. Trails [Stable Code 3B](stable-code-3b.md) here, though they rank lower overall. 
test   result  vs the others in its run  difficulty      counts  why
GSM8K   28.0%  ▆▂▅●▅▅▃▆···▂ beats 8/25   easy · avg 72%    2.5%  sharp · at its level
other tests (9) 
test                                result  vs the others in its run    difficulty        counts  why
LiveBench 2024-07-26                 16.7%  ▁▁▁▂●▃▂█▆▃▅▃ beats 10/61    medium · avg 38%    8.6%  typical
BIG-Bench Hard                       44.9%  ▁·····▂●▂▅·▁ beats 5/25     easy · avg 74%      4.4%  sharp
MMLU                                 58.6%  ▁▁▃▃▄█●▁█··▁ beats 17/25    medium · avg 73%    3.9%  at its level · several runs
Natural Questions (closed book)      25.2%  ▁·▁·▁▂▁▂·▅▂● beats 20/28    medium · avg 48%    3.6%  typical
Open PT-LLM Leaderboard average (…   64.5%  ▃▁▁▁▁▁▁▁▂●▅▁ beats 111/276  medium · avg 72%    0.8%  at its level · noisy
Tulu 2: TyDiQA gold passage, 1-sh…   23.7%  ●··▁▃▂▂▂█▅▁▂ beats 1/25     medium · avg 59%    0.6%  blunt · at its level
TriviaQA                             64.4%  ▁··▁█▃▃▁▄▅▄● beats 24/28    easy · avg 73%      0.5%  at its level · noisy
HaluEval QA hallucination detecti…   51.8%  █▁▁·●▁·▁▁▁▁▁ beats 19/27    hard · avg 66%      0.3%  noisy
TruthfulQA MC2                       55.1%  ▅▃█▆▁▂●···▁▁ beats 26/31    medium · avg 57%    0.1%  blunt · at its level
## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's 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 is added up by who ran the tests. [Method](../method.md) 
source                                    results  share of the score
EvalPlus / EvalPerf published evaluatio…        3   54% ▇▇▇▇▇▇
Tulu 2                                          5   12% ▇
LiveBench model judgments                       1    9% ▇
AlpacaEval 1.0 and 2.0 leaderboards             2    7% ▇
OpenChat 3.5 model card                         2    5% ▇
Hallucinations Leaderboard results              4    4% ▇
LMArena leaderboard dataset                     1    4% ▇
Polyrating: A Cost-Effective and Bias-A…        1    4% ▇
Open Portuguese LLM Leaderboard                 1    1% ▇
prior: pull toward the middle                        0% ▇
weighted average                                   = 35.4
Licences: EvalPlus / EvalPerf published evaluation runs ([Apache-2.0](../../licenses/evalplus-published-runs-Apache-2.0.txt)) · Tulu 2 ([CC BY 4.0](../../licenses/CC-BY-4.0.txt)) · LiveBench model judgments ([Apache-2.0 (stated in the LiveBench datasheet)](https://www.apache.org/licenses/LICENSE-2.0)) · AlpacaEval 1.0 and 2.0 leaderboards ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)) · OpenChat 3.5 model card ([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/)) · Open Portuguese LLM Leaderboard ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt))
Not counted (2) 
source                  test            score  why
OpenChat 3.5 model ca…  BIG-Bench Hard  16.0%  failed run: Few-shot CoT format failure in a chat-mode harness: 16.0 vs 44.9 for the same checkpoint in the Tulu 2 paper (3-shot CoT), and below BBH's published random average of 25.7.
OpenChat 3.5 model ca…  GSM8K            5.1%  failed run: Same few-shot CoT format failure: 5.1 vs 28.0 in the Tulu 2 paper's 8-shot CoT run.
## 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 Zephyr 7B beta (Q4_K_M, needs ~4.4 GB, from [huggingface.co/TheBloke](https://huggingface.co/TheBloke/zephyr-7B-beta-GGUF/tree/e4714d14e9652aa9658fa937732cceadc63ac42e)) ```sh llama cli -hf TheBloke/zephyr-7B-beta-GGUF:Q4_K_M ``` 3. Or start a local server and open the address it prints (http://localhost:9931) ```sh llama serve -hf TheBloke/zephyr-7B-beta-GGUF:Q4_K_M ``` Other versions: Q8_0 (~7.7 GB) `llama cli -hf TheBloke/zephyr-7B-beta-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 Zephyr 7B beta (Q4_K_M, needs ~4.4 GB, from [huggingface.co/TheBloke](https://huggingface.co/TheBloke/zephyr-7B-beta-GGUF/tree/e4714d14e9652aa9658fa937732cceadc63ac42e)) ```sh lms get https://huggingface.co/TheBloke/zephyr-7B-beta-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 (~7.7 GB) `lms get https://huggingface.co/TheBloke/zephyr-7B-beta-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 Zephyr 7B beta (Q4_K_M, needs ~4.84 GB, from [ollama.com](https://ollama.com/library/zephyr/tags)) ```sh ollama run zephyr:7b-beta-q4_K_M ``` Other versions: Q8_0 (~9.19 GB) `ollama run zephyr:7b-beta-q8_0` Model repository and documentation: [huggingface.co/HuggingFaceH4](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) [interactive version](https://whatmodel.sh/model/zephyr-7b-beta) · [JSON](https://whatmodel.sh/api?format=json&all=1)