[whatmodel.sh](../../README.md) / OpenThinker2 32B
# OpenThinker2 32B
Open Thoughts · released 2025-04 · 32.76B params · apache-2.0
**score 54.6** ±1.4 · #293 of 732 · from 6 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 OpenThinker2 32B among every model on that task or test, worst to best.
task vs the others rank counts tests
chat & writing ▁▁▁▃▆▄●█▆▇▇▃ #273 of 402* 22% 2
math ▁▁▃▅█▆●█▄▂▁▁ #171 of 431 48% 3
chat & writing · #273 of 402 · 22% of its score
Beats [Ministral 3 14B Reasoning 2512](ministral-3-14b-reasoning.md), [Ministral 3 14B](ministral-3-14b.md) here, though they rank higher overall. Trails [Flan-T5 XXL](flan-t5-xxl.md), [North Micro Vision Instruct (2.4B)](north-micro-vision-instruct.md), [Granite 4.0 H 1B](granite-4-0-h-1b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
Arena-Hard v2.0 hard prompts (sty… 3.2% ●·▃▅▅·▃█▆▂▂▂ beats 0/22 very hard · avg 12% 17% sharp
Arena-Hard v2.0 creative writing … 23.6% ▆▂▂●▂▂▃█▂▃·▂ beats 6/22 hard · avg 19% 4.2% sharp · at its level · noisy
math · #171 of 431 · 48% of its score
Beats [MiMo-V2.5 (310B-A15B)](mimo-v2-5.md), [MiniMax-M2.7](minimax-m2-7.md), [MiniMax-M2.1](minimax-m2-1.md) here, though they rank higher overall. Trails [OpenThinker3 7B](openthinker3-7b.md), [Olmo 3 7B Think](olmo-3-7b-think.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
AIME 2025 (maker) 58.7% ▁·▁▁▁▂▅●█▇▅▁ beats 45/133 medium · avg 37% 20% sharp · at its level
AIME 2024 (maker) 76.7% ▁▁▁▂▁▃▁▅█●█▃ beats 47/85 medium · avg 56% 17% sharp · at its level
MATH-500 (maker) 90.8% ▁▃▂▃▄▇●▇█▄▂▁ beats 42/85 very easy · avg 86% 11% typical
other tests (1)
test result vs the others in its run difficulty counts why
GPQA Diamond (maker) 64.1% ▁▁▁▁▂▄▅●███▄ beats 128/322 hard · avg 51% 29% sharp · at its level
source results share of the score
open-thoughts/OpenThinker2-32B model ca… 4 78% ▇▇▇▇▇▇▇▇▇
Arena-Hard-Auto v2.0-Preview leaderboard 2 22% ▇▇▇
prior: pull toward the middle 1% ▇
weighted average = 54.6
Licences: open-thoughts/OpenThinker2-32B model card ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)) · Arena-Hard-Auto v2.0-Preview leaderboard ([Apache-2.0](../../licenses/arena-hard-auto-Apache-2.0.txt))
## 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 OpenThinker2 32B (Q4_K_M, needs ~19.9 GB, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/OpenThinker2-32B-GGUF/tree/4e404631501e91ff8e25818db98bfcea9f1b42b8))
```sh
llama cli -hf lmstudio-community/OpenThinker2-32B-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf lmstudio-community/OpenThinker2-32B-GGUF:Q4_K_M
```
Other versions: Q8_0 (~34.8 GB) `llama cli -hf lmstudio-community/OpenThinker2-32B-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 OpenThinker2 32B (Q4_K_M, needs ~19.9 GB, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/OpenThinker2-32B-GGUF/tree/4e404631501e91ff8e25818db98bfcea9f1b42b8))
```sh
lms get https://huggingface.co/lmstudio-community/OpenThinker2-32B-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 (~34.8 GB) `lms get https://huggingface.co/lmstudio-community/OpenThinker2-32B-GGUF@Q8_0 -y` · 4bit (~18.4 GB, Mac only) `lms get https://huggingface.co/alexgusevski/OpenThinker2-32B-mlx-4Bit -y` · 8bit (~34.8 GB, Mac only) `lms get https://huggingface.co/alexgusevski/OpenThinker2-32B-mlx-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 OpenThinker2 32B (Q4_K_M, needs ~19.9 GB, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/OpenThinker2-32B-GGUF/tree/4e404631501e91ff8e25818db98bfcea9f1b42b8))
```sh
ollama run hf.co/lmstudio-community/OpenThinker2-32B-GGUF:Q4_K_M
```
Other versions: Q8_0 (~34.8 GB) `ollama run hf.co/lmstudio-community/OpenThinker2-32B-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 OpenThinker2 32B (4bit, needs ~18.4 GB, Mac only, from [huggingface.co/alexgusevski](https://huggingface.co/alexgusevski/OpenThinker2-32B-mlx-4Bit/tree/98ccf4eb104298284163ff2e7d652039e3f41570))
```sh
mlx_lm.chat --model alexgusevski/OpenThinker2-32B-mlx-4Bit
```
3. Or start an OpenAI-compatible server at http://localhost:8080/v1
```sh
mlx_lm.server --model alexgusevski/OpenThinker2-32B-mlx-4Bit
```
Other versions: 8bit (~34.8 GB, Mac only) `mlx_lm.chat --model alexgusevski/OpenThinker2-32B-mlx-8Bit`
### Other ways to run it
1. Run it as published (BF16, needs ~65.53 GB, NVIDIA only, from [huggingface.co/open-thoughts](https://huggingface.co/open-thoughts/OpenThinker2-32B/tree/0aefafe3ee27e073e14fb94a0563795fddf38c50))
```sh
vllm serve open-thoughts/OpenThinker2-32B
```
Model repository and documentation: [huggingface.co/open-thoughts](https://huggingface.co/open-thoughts/OpenThinker2-32B)
[interactive version](https://whatmodel.sh/model/openthinker2-32b) · [JSON](https://whatmodel.sh/api?format=json&all=1)