[whatmodel.sh](../../README.md) / Qwen2-VL 7B Instruct
# Qwen2-VL 7B Instruct
Alibaba · released 2024-08 · 8.29B params · apache-2.0
**score 46.6** ±0.9 · #444 of 732 · from 27 benchmark results across 6 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks Qwen2-VL 7B Instruct among every model on that task or test, worst to best.
task vs the others rank counts tests
coding ▁▁▁▃▅●█▇▅▄▃▁ #329 of 469* 2.3% 1
math ▁▁▃●█▆▄█▄▂▁▁ #354 of 431 19% 3
vision & documents ▁▁▂●▁▃▄▂▆▇█▄ #133 of 153 79% 17
coding · #329 of 469 · 2.3% of its score
Beats [Qwen3-VL 8B Instruct](qwen3-vl-8b-instruct.md), [Qwen3-VL 4B Instruct](qwen3-vl-4b-instruct.md), [LFM2.5 2.6B](lfm2-5-2-6b.md) here, though they rank higher overall. Trails [DeepSeek LLM 67B Chat](deepseek-llm-67b-chat.md), [WizardLM 70B V1.0](wizardlm-70b-v1-0.md), [Llama 3.2 11B Vision Instruct](llama-3-2-11b-vision.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
HumanEval 62.2% ▁▁▁▁▂▂▆▆●▅▂▁ beats 117/205 easy · avg 67% 2.3% sharp · at its level · noisy
math · #354 of 431 · 19% of its score
Beats [GPT-3.5 Turbo (0125)](gpt-3-5-turbo-0125.md), [DBRX Instruct](dbrx-instruct.md), [Phi-3 Medium 128K Instruct](phi-3-medium-128k-instruct.md) here, though they rank higher overall. Trails [OPT-2.7B](opt-2-7b.md), [Gemma 2B (base)](gemma-2b-base.md), [Stable Code 3B](stable-code-3b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
MathVista (testmini) (maker) 58.2% ▃··●▅·▃█▅▅▅█ beats 1/16 medium · avg 69% 12% at its level · several runs
MATH-Vision (maker) 16.3% ▁●·▁··▄▁▃▇▇█ beats 1/24 hard · avg 25% 3.8% small run
MATH (subset unspecified) 27.9% ▂▄●▃█▇▅▆▄▃▄▁ beats 11/80 medium · avg 51% 2.7% at its level
vision & documents · #133 of 153 · 79% of its score
Beats [Nova Pro](nova-pro.md), [Nova Lite](nova-lite.md), [Gemma 4 E2B](gemma-4-e2b.md) here, though they rank higher overall.
test result vs the others in its run difficulty counts why
MMMU (val) (maker) 54.1% ▂···●▂▂▂▄▄▄█ beats 3/16 medium · avg 61% 15% at its level · several runs
MathVista (testmini) (maker) 58.2% ▃··●▅·▃█▅▅▅█ beats 1/16 medium · avg 69% 12% at its level · several runs
MMStar (maker) 60.7% ▂··▄●▂·▄█▂▄▆ beats 3/16 medium · avg 65% 8.9% at its level
MMMU Pro (maker) 43.5% ▃·█·●▃▃▃▅▅▃█ beats 5/16 medium · avg 46% 6.4% at its level
Arena vision (style controlled) 1030.6 ▁▁▂▂●▂▃▄▅▅█▇ beats 18/152 avg 1061 5.9% blunt
MATH-Vision (maker) 16.3% ▁●·▁··▄▁▃▇▇█ beats 1/24 hard · avg 25% 3.8% small run
CapArena-Auto detailed image capt… -49.0000 ▃●▃·▃█·▅▃▃▃▃ beats 1/12 avg -4.498 3.7% typical
HallusionBench (maker) 50.6% ▃·▃·●▃▅▅▃▅█▅ beats 2/16 medium · avg 54% 3.6% blunt · at its level
ChartQA (maker) 83.0% ▂··▂·●▃·▅▂▆█ beats 2/17 easy · avg 85% 3.5% blunt · several runs
DocVQA (maker) 94.5% ▃··▅▃●▃█▃█▅▅ beats 5/17 very easy · avg 94% 3.2% blunt · maxed out
InfographicVQA (maker) 76.5% ▃·▃▅█●▃▃▃▃▃▃ beats 7/13 easy · avg 80% 3.0% small run
RealWorldQA (maker) 70.1% ▅▃·▃●▅·█·▃▅█ beats 5/16 easy · avg 70% 2.8% blunt · at its level
MM-MT-Bench (Mistral, GPT-4o judg… 5.4500 MM-MT-Bench score (1-10) ▅··▃·●·▃▃··▅ beats 3/9 avg 5.605 MM-MT-Bench s… 2.7% sharp · noisy
VQAv2 (val) 75.9% ▃···█▃··●▃▃▃ beats 6/9 easy · avg 73% 1.6% blunt · at its level
OCRBench (maker) 86.6% ▅█·▃▃●▃▅▅▃·▃ beats 9/16 easy · avg 84% 1.1% blunt
Ai2 Molmo human preference Elo (S… 1025.0 ▄▄·▄·▄●▄·▄·█ beats 4/8 avg 1032 0.9% blunt
TextVQA (maker) 84.3% █······▄·●▄▄ beats 3/5 easy · avg 80% 0.6% blunt
other tests (5)
test result vs the others in its run difficulty counts why
MMLU 68.5% ▁▁▁▁▁▂▂▅█●▆▄ beats 127/214 easy · avg 75% 6.0% at its level
Video-MME (without subtitles) (ma… 63.3% ▃··●··▅█▃▃▅▃ beats 3/13 medium · avg 66% 4.9% blunt · at its level
Video-MME (with subtitles) (maker) 69.0% ▁▁▁▁●·▁▃▁▁▁█ beats 4/16 medium · avg 73% 2.1% at its level · small run
MVBench (maker) 67.0% ▂▂▂·●█··▆▄·▂ beats 3/13 medium · avg 67% 2.0% blunt · at its level
Vectara hallucination leaderboard… 95.8% ▁▁▂▃▄▄●▆▃▂▂▁ beats 82/140 very easy · avg 94% 0.7% blunt · maxed out · noisy
source results share of the score
Qwen (submitted / maker-reported) 15 70% ▇▇▇▇▇▇▇▇
Pixtral 12B 8 19% ▇▇
LMArena leaderboard dataset 1 6% ▇
CapArena: Benchmarking and Analyzing De… 1 4% ▇
Ai2 Molmo-72B-0924 model card: Human Pr… 1 1% ▇
Vectara hallucination leaderboard 1 1% ▇
prior: pull toward the middle 0% ▇
weighted average = 46.6
Licences: Pixtral 12B ([CC BY 4.0](http://creativecommons.org/licenses/by/4.0/)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · CapArena: Benchmarking and Analyzing Detailed Image Captioning in the LLM Era ([CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)) · Ai2 Molmo-72B-0924 model card: Human Preference Elo Rating column ([Apache-2.0](https://huggingface.co/allenai/Molmo-72B-0924/blob/ab586dd9640de059566975b8e5bcd0c6fdc73956/README.md)) · Vectara hallucination leaderboard ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0))
## 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 Qwen2-VL 7B Instruct (Q4_K_M, needs ~5.4 GB, from [huggingface.co/ggml-org](https://huggingface.co/ggml-org/Qwen2-VL-7B-Instruct-GGUF/tree/237a3f007d71a31ea56e4a9c1a11f083e7c1cba6))
```sh
llama cli -hf ggml-org/Qwen2-VL-7B-Instruct-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf ggml-org/Qwen2-VL-7B-Instruct-GGUF:Q4_K_M
```
Other versions: Q8_0 (~8.82 GB) `llama cli -hf ggml-org/Qwen2-VL-7B-Instruct-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 Qwen2-VL 7B Instruct (Q4_K_M, needs ~5.4 GB, from [huggingface.co/ggml-org](https://huggingface.co/ggml-org/Qwen2-VL-7B-Instruct-GGUF/tree/237a3f007d71a31ea56e4a9c1a11f083e7c1cba6))
```sh
lms get https://huggingface.co/ggml-org/Qwen2-VL-7B-Instruct-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 (~8.82 GB) `lms get https://huggingface.co/ggml-org/Qwen2-VL-7B-Instruct-GGUF@Q8_0 -y` · 4bit (~4.7 GB, Mac only) `lms get https://huggingface.co/mlx-community/Qwen2-VL-7B-Instruct-4bit -y` · 8bit (~8.8 GB, Mac only) `lms get https://huggingface.co/mlx-community/Qwen2-VL-7B-Instruct-8bit -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 Qwen2-VL 7B Instruct (Q4_K_M, needs ~5.4 GB, from [huggingface.co/ggml-org](https://huggingface.co/ggml-org/Qwen2-VL-7B-Instruct-GGUF/tree/237a3f007d71a31ea56e4a9c1a11f083e7c1cba6))
```sh
ollama run hf.co/ggml-org/Qwen2-VL-7B-Instruct-GGUF:Q4_K_M
```
Other versions: Q8_0 (~8.82 GB) `ollama run hf.co/ggml-org/Qwen2-VL-7B-Instruct-GGUF:Q8_0`
### MLX (Mac)
1. Install MLX (Mac). Apple silicon Macs only, text only.
```sh
# with uv
uv tool install mlx-lm
# or
pip install mlx-lm
```
2. Run Qwen2-VL 7B Instruct (4bit, needs ~4.7 GB, Mac only, from [huggingface.co/mlx-community](https://huggingface.co/mlx-community/Qwen2-VL-7B-Instruct-4bit/tree/1c638e970be36948e087e77d103b9147e1a9a1dd))
```sh
mlx_lm.chat --model mlx-community/Qwen2-VL-7B-Instruct-4bit
```
3. Or start an OpenAI-compatible server at http://localhost:8080/v1
```sh
mlx_lm.server --model mlx-community/Qwen2-VL-7B-Instruct-4bit
```
Other versions: 8bit (~8.8 GB, Mac only) `mlx_lm.chat --model mlx-community/Qwen2-VL-7B-Instruct-8bit`
Model repository and documentation: [huggingface.co/Qwen](https://huggingface.co/Qwen/Qwen2-VL-7B-Instruct)
[interactive version](https://whatmodel.sh/model/qwen2-vl-7b) · [JSON](https://whatmodel.sh/api?format=json&all=1)