[whatmodel.sh](../../README.md) / Qwen2-VL 2B Instruct
# Qwen2-VL 2B Instruct
Alibaba · released 2024-08 · 2.21B params · apache-2.0
**score 38.1** ±1.1 · #595 of 732 · from 16 benchmark results across 2 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 2B Instruct among every model on that task or test, worst to best.
task vs the others rank counts tests
math ▁▁●▅█▆▄█▄▂▁▁ #409 of 431* 19% 2
vision & documents ●▁▂▂▁▃▄▂▆▇█▄ #151 of 153 85% 12
math · #409 of 431 · 19% of its score
Beats [Command R (03-2024)](command-r-03-2024.md), [Mistral 7B Instruct v0.3](mistral-7b-instruct-v0-3.md) here, though they rank higher overall. Trails [GPT-Neo 125M](gpt-neo-125m.md), [OPT-125M](opt-125m.md), [GPT-2 XL (1.5B)](gpt-2-xl.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
MathVista (testmini) (maker) 43.0% ●··▃▅·▃█▅▅▅█ beats 0/16 medium · avg 69% 15% at its level
MATH-Vision (maker) 12.4% ●▁·▁··▄▁▃▇▇█ beats 0/24 hard · avg 25% 3.7% too hard for it · small run
vision & documents · #151 of 153 · 85% of its score
Beats [Gemma 3 4B](gemma-3-4b.md), [Phi-3-vision-128k-instruct](phi-3-vision-128k-instruct.md) here, though they rank higher overall. Trails [Qwen3.5 0.8B](qwen3-5-0-8b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
MMMU (val) (maker) 41.1% ●···▆▂▂▂▄▄▄█ beats 0/16 medium · avg 61% 16% at its level
MathVista (testmini) (maker) 43.0% ●··▃▅·▃█▅▅▅█ beats 0/16 medium · avg 69% 15% at its level
MMStar (maker) 48.0% ●··▄▂▂·▄█▂▄▆ beats 0/16 medium · avg 65% 12% at its level
MMMU Pro (maker) 37.6% ▃·●·▅▃▃▃▅▅▃█ beats 2/16 medium · avg 46% 7.1% typical
DocVQA (maker) 90.1% ●··▅▃▅▃█▃█▅▅ beats 0/17 very easy · avg 94% 6.5% blunt
ChartQA (maker) 73.5% ●··▂·▂▃·▅▂▆█ beats 0/17 easy · avg 85% 6.2% blunt
HallusionBench (maker) 41.7% ●·▃·▅▃▅▅▃▅█▅ beats 0/16 medium · avg 54% 5.3% blunt · at its level
InfographicVQA (maker) 65.5% ▃·▃●█▃▃▃▃▃▃▃ beats 3/13 easy · avg 80% 5.2% at its level · small run
RealWorldQA (maker) 62.9% ●▃·▃▅▅·█·▃▅█ beats 0/16 easy · avg 70% 4.6% blunt · at its level
MATH-Vision (maker) 12.4% ●▁·▁··▄▁▃▇▇█ beats 0/24 hard · avg 25% 3.7% too hard for it · small run
OCRBench (maker) 80.9% ▅●·▃▃█▃▅▅▃·▃ beats 3/16 easy · avg 84% 1.9% blunt
TextVQA (maker) 79.7% ●······▄·▄▄▄ beats 1/5 easy · avg 80% 0.9% blunt
other tests (4)
test result vs the others in its run difficulty counts why
Video-MME (without subtitles) (ma… 55.6% ●··█··▅█▃▃▅▃ beats 0/13 medium · avg 66% 7.1% blunt · at its level
Video-MME (with subtitles) (maker) 60.4% ▁▁●▁▁·▁▃▁▁▁█ beats 2/16 medium · avg 73% 3.3% at its level · small run
MVBench (maker) 63.2% ▂●▂·▂█··▆▄·▂ beats 1/13 medium · avg 67% 3.0% blunt · at its level
Vectara hallucination leaderboard… 91.7% ▁▁▂▃●▄█▆▃▂▂▁ beats 33/140 very easy · avg 94% 1.4% blunt · noisy
source results share of the score
Qwen (submitted / maker-reported) 15 98% ▇▇▇▇▇▇▇▇▇▇▇▇
Vectara hallucination leaderboard 1 1% ▇
prior: pull toward the middle 1% ▇
weighted average = 38.1
Licences: 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 2B Instruct (Q4_K_M, needs ~1.7 GB, from [huggingface.co/ggml-org](https://huggingface.co/ggml-org/Qwen2-VL-2B-Instruct-GGUF/tree/bb307c036e8a1ed7b663bbd0c35b41c4c9294cfd))
```sh
llama cli -hf ggml-org/Qwen2-VL-2B-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-2B-Instruct-GGUF:Q4_K_M
```
### 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 2B Instruct (Q4_K_M, needs ~1.7 GB, from [huggingface.co/ggml-org](https://huggingface.co/ggml-org/Qwen2-VL-2B-Instruct-GGUF/tree/bb307c036e8a1ed7b663bbd0c35b41c4c9294cfd))
```sh
lms get https://huggingface.co/ggml-org/Qwen2-VL-2B-Instruct-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: 4bit (~1.2 GB, Mac only) `lms get https://huggingface.co/mlx-community/Qwen2-VL-2B-Instruct-4bit -y` · 8bit (~2.3 GB, Mac only) `lms get https://huggingface.co/mlx-community/Qwen2-VL-2B-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 2B Instruct (Q4_K_M, needs ~1.7 GB, from [huggingface.co/ggml-org](https://huggingface.co/ggml-org/Qwen2-VL-2B-Instruct-GGUF/tree/bb307c036e8a1ed7b663bbd0c35b41c4c9294cfd))
```sh
ollama run hf.co/ggml-org/Qwen2-VL-2B-Instruct-GGUF:Q4_K_M
```
### 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 2B Instruct (4bit, needs ~1.2 GB, Mac only, from [huggingface.co/mlx-community](https://huggingface.co/mlx-community/Qwen2-VL-2B-Instruct-4bit/tree/01af461cdb9574acc09084a0ef94e216e142b085))
```sh
mlx_lm.chat --model mlx-community/Qwen2-VL-2B-Instruct-4bit
```
3. Or start an OpenAI-compatible server at http://localhost:8080/v1
```sh
mlx_lm.server --model mlx-community/Qwen2-VL-2B-Instruct-4bit
```
Other versions: 8bit (~2.3 GB, Mac only) `mlx_lm.chat --model mlx-community/Qwen2-VL-2B-Instruct-8bit`
Model repository and documentation: [huggingface.co/Qwen](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct)
[interactive version](https://whatmodel.sh/model/qwen2-vl-2b) · [JSON](https://whatmodel.sh/api?format=json&all=1)