Instructions to use MiniMaxAI/SynLogic-32B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MiniMaxAI/SynLogic-32B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="MiniMaxAI/SynLogic-32B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MiniMaxAI/SynLogic-32B")
model = AutoModelForCausalLM.from_pretrained("MiniMaxAI/SynLogic-32B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use MiniMaxAI/SynLogic-32B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MiniMaxAI/SynLogic-32B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MiniMaxAI/SynLogic-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/MiniMaxAI/SynLogic-32B

SGLang

How to use MiniMaxAI/SynLogic-32B with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "MiniMaxAI/SynLogic-32B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MiniMaxAI/SynLogic-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "MiniMaxAI/SynLogic-32B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MiniMaxAI/SynLogic-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use MiniMaxAI/SynLogic-32B with Docker Model Runner:
```
docker model run hf.co/MiniMaxAI/SynLogic-32B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

SynLogic-32B: Advanced Logical Reasoning Model

🐙 GitHub Repo: https://github.com/MiniMax-AI/SynLogic
📜 Paper (arXiv): https://arxiv.org/abs/2505.19641
🤗 Dataset: SynLogic on Hugging Face

Model Overview

SynLogic-32B is a state-of-the-art reasoning model built on Qwen2.5-32B-Base and trained using reinforcement learning on our comprehensive SynLogic dataset. The model excels at logical reasoning tasks and demonstrates strong generalization to mathematical domains.

Key Features

Comprehensive Logical Reasoning: Trained on 35 diverse logical reasoning tasks including Sudoku, Game of 24, Cipher, Arrow Maze, and more
Verifiable Training: All training data can be automatically verified, enabling effective reinforcement learning
Strong Generalization: Transfers logical reasoning skills to mathematical problem-solving without explicit math training

Performance Highlights

Model	BBEH	KOR-Bench	BBH
Qwen2.5-32B-Instruct	17.5	54.7	84.5
DeepSeek-R1-Distill-Qwen-32B	19.2	66.6	88.3
SynLogic-32B	25.5	62.2	85.8

Key Achievement: +6 points improvement over DeepSeek-R1-Distill-Qwen-32B on the challenging BBEH benchmark, establishing state-of-the-art performance among open-source logical reasoning models.

Training Details

Base Model: Qwen2.5-32B-Base
Training Algorithm: GRPO (Group Relative Policy Optimization)
Dataset: 33k SynLogic-Hard samples with controlled difficulty
Reward Design: Binary rewards based on format adherence and correctness verification

Citation

@misc{liu2025synlogic,
      title={SynLogic: Synthesizing Verifiable Reasoning Data at Scale for Learning Logical Reasoning and Beyond}, 
      author={Junteng Liu and Yuanxiang Fan and Zhuo Jiang and Han Ding and Yongyi Hu and Chi Zhang and Yiqi Shi and Shitong Weng and Aili Chen and Shiqi Chen and Yunan Huang and Mozhi Zhang and Pengyu Zhao and Junjie Yan and Junxian He},
      year={2025},
      eprint={2505.19641},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2505.19641}, 
}