whatmodel.sh / Qwen2.5-VL 32B Instruct

Qwen2.5-VL 32B Instruct

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

Qwen2.5-VL 32B Instruct is an open-weights language model from Alibaba, released March 2025. Its published size is 33.5B parameters and a 33k-token context window. Licence: apache-2.0.

Catalogue identity notes

Arena entry 'qwen2.5-vl-32b-instruct' (vision). LMArena's leaderboard table (arena-leaderboard space @6f2eeeb) labels 'qwen2.5-vl-32b-instruct' as 'Qwen2.5-VL-32B-Instruct' (Alibaba, Apache 2.0), linking https://huggingface.co/Qwen/Qwen2.5-VL-32B-Instruct. Maker repo Qwen/Qwen2.5-VL-32B-Instruct @7cfb30d: licence tag apache-2.0. 33,452,718,336 parameters from the Hugging Face safetensors metadata (dense, so active = total). Released 2025-03: month the maker's Hugging Face repo was created (createdAt 2025-03-21).

53.0±1.5#297 of 719 LLMsfrom 16 benchmark results across 7 sources

$ how it compares

The LLMs ranked either side of Qwen2.5-VL 32B Instruct, out of 828 with a score, then more from Alibaba.

#modelscorewhereRAMreleased
355Llama 3.1 Tulu 3 70B DPO Ai253±1local~22GB2024-11
356Qwen Turbo (2024-11-01) Alibaba53±2comparison only2024-11
357Qwen2.5-VL 32B Instruct Alibaba53±2local~22GB2025-03
358DeepSeek-Coder-V2 Instruct (0614) DeepSeek53±1local~92GB2024-06
359s1.1 32B Simple Scaling53±1local~22GB2025-02
more from Alibaba
18Qwen3.8 Max (2.4T-A95B) Alibaba78±1api $3.00/M +local~380GB2026-08
34Qwen3.8 Max (0902) Alibaba76±1api $3.00/M2026-09
37Qwen3.8-Flash-Next Alibaba75±1local~59GB2026-08

see all LLMs · find the ones that fit your machine

$ what it's good at

vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #110 of 147 · 77% of its score · 11 tests

Beats DeepSeek V4 Flash Vision (Exp), GPT-4 Turbo (2024-04-09), Pixtral Large (2411) here, though they rank higher overall. Trails Gemma 3 27B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena vision (style controlled)1118.6 ✎▁▁▂▂▂▂▄▅▅▆██ beats 34/145avg 1068▇▇▇▇▇▇ 14%blunt
ChartArena chart parsing (English, mAP_high average)31.8% ✎▁·▁·▃▃▃▇▃█·▄ beats 6/22hard · avg 27%▇▇▇▇ 10%blunt · at its level
CharXiv (reasoning questions)39.1% ✎▂▅▃▅▅▆▆▅▆█▆▂ beats 16/100medium · avg 38%▇▇▇▇ 9.2%at its level · small run
CharXiv descriptive questions72.5% ✎▂▄▄▅▇▄█▇▇▂▁▂ beats 16/58easy · avg 72%▇▇▇▇ 9.0%small run
Kaleidoscope (in-language multimodal exams, 18 languages, overall accuracy)48.2% ✎█·█···█·█·██ beats 2/5medium · avg 47%▇▇▇ 8.3%at its level
MMMU Pro lab49.5% ✎▂▃▁▃▆▇▃▇▇██▅ beats 10/52medium · avg 44%▇▇▇ 7.8%at its level · small run
MMStar lab69.5% ✎▃▅▂▃▂▂▆█▆▆█▅ beats 12/34medium · avg 64%▇▇ 6.2%at its level · small run
VMMU (Vietnamese multimodal understanding and reasoning, mean accuracy over 7 tasks)52.1% ✎▅······▃·▅·█ beats 7/7very hard · avg 36%▇▇ 4.7%noisy
3 more tests
testresultvs the others in its rundifficultycountswhy
ScreenSpot-Pro lab39.4% ✎▂▃▅▆██▅▂···▂ beats 1/24medium · avg 44%▇ 3.0%blunt · at its level · small run
DocVQA lab94.8% ✎▁··▃▁▁▁▇█▇▄▅ beats 15/28very easy · avg 94%▇ 2.7%blunt · maxed out · small run
CC-OCR lab77.1% ✎▁·▂··▂▂▃▃▂▆█ beats 10/26easy · avg 72%▇ 1.7%blunt · at its level · small run
other tests5 not in a task · 22% of its score
testresultvs the others in its rundifficultycountswhy
Video-MME (with subtitles) lab77.9% ✎▁▁▁▁▁·▁▃▁▁▁█ beats 7/16medium · avg 73%▇▇ 5.9%at its level · small run
LVBench lab49.0% ✎▇·█▅▅▄▄··▃▁▁ beats 4/28hard · avg 49%▇▇ 4.4%at its level · small run
Open PT-LLM Leaderboard average (9 Portuguese tasks)80.9% ✎▃▁▁▁▁▁▁▁▂█▅▁ beats 263/276medium · avg 73%▇▇ 4.2%at its level · noisy
Video-MME (without subtitles) lab70.5% ✎▁▃▁▃▁·█▄▄▅·▇ beats 10/27medium · avg 67%▇▇ 4.1%blunt · at its level · small run
AndroidWorld lab22.0% ✎▃·▁▁▂▃█▅▂▂·▂ beats 0/24hard · avg 34%▇ 3.6%blunt · at its level · small run

$ where to run it

llama.cpp · recommended

version

  1. 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 ↗

  2. Run Qwen2.5-VL 32B Instruct · needs ~19.9 GB · from huggingface.co/ggml-org
    chat in the terminal
    llama cli -hf ggml-org/Qwen2.5-VL-32B-Instruct-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf ggml-org/Qwen2.5-VL-32B-Instruct-GGUF:Q4_K_M

LM Studio

version

  1. Install LM Studio · Open the app once before using the lms command.

    download it (macOS / Windows / Linux) ↗

  2. Run Qwen2.5-VL 32B Instruct · needs ~19.9 GB · from huggingface.co/ggml-org

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/ggml-org/Qwen2.5-VL-32B-Instruct-GGUF@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

version

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run Qwen2.5-VL 32B Instruct · needs ~20.6 GB · from ollama.com
    chat in the terminal
    ollama run qwen2.5vl:32b-q4_K_M

MLX (Mac)

version

  1. Install MLX (Mac) · Apple silicon Macs only, text only.
    with uv
    uv tool install mlx-lm
    or
    pip install mlx-lm
  2. Run Qwen2.5-VL 32B Instruct · needs ~19.8 GB, Mac only · from huggingface.co/mlx-community
    chat in the terminal
    mlx_lm.chat --model mlx-community/Qwen2.5-VL-32B-Instruct-4bit
  3. Or start an OpenAI-compatible server at http://localhost:8080/v1
    mlx_lm.server --model mlx-community/Qwen2.5-VL-32B-Instruct-4bit

Official weights and documentation: huggingface.co/Qwen ↗

$ how the score was worked out · Qwen2.5-VL-32B-Instruct model… 39%, Apriel-1.5-15b-Thinker (arXiv… 18%, LMArena leaderboard dataset 14% and 4 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
Qwen2.5-VL-32B-Instruct model card · lab-reported9▇▇▇▇▇▇▇▇▇▇▇▇ 39%Apache-2.0
Apriel-1.5-15b-Thinker (arXiv:2510.01141v1), Table 4 multimodal benchmarks2▇▇▇▇▇ 18%CC-BY-4.0
LMArena leaderboard dataset1▇▇▇▇ 14%CC BY 4.0
ChartArena: Benchmarking Chart Parsing across Languages, Scenarios, and Formats (arXiv:2606.01348v3)1▇▇▇ 10%CC BY 4.0
Kaleidoscope: In-language Exams for Massively Multilingual Vision Evaluation (arXiv:2504.07072v2)1▇▇▇ 8%CC-BY-4.0
VMMU: A Vietnamese Multitask Multimodal Understanding and Reasoning Benchmark (arXiv:2508.13680v5)1▇ 5%CC-BY-4.0
Open Portuguese LLM Leaderboard (archived)1▇ 4%Apache-2.0
prior · pull toward the middle▇ 1%
weighted average= 53.0

$ 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