Instructions to use MiniMaxAI/MiniMax-M2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MiniMaxAI/MiniMax-M2 with Transformers:

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

pipe = pipeline("text-generation", model="MiniMaxAI/MiniMax-M2", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MiniMaxAI/MiniMax-M2", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("MiniMaxAI/MiniMax-M2", trust_remote_code=True, device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use MiniMaxAI/MiniMax-M2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MiniMaxAI/MiniMax-M2"
# 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/MiniMax-M2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/MiniMaxAI/MiniMax-M2

SGLang

How to use MiniMaxAI/MiniMax-M2 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/MiniMax-M2" \
    --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/MiniMax-M2",
		"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/MiniMax-M2" \
        --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/MiniMax-M2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use MiniMaxAI/MiniMax-M2 with Docker Model Runner:
```
docker model run hf.co/MiniMaxAI/MiniMax-M2
```

yzhuang commited on 7 days ago

Commit

e3436cb

verified ·

1 Parent(s): 04312f0

Add PLE (Physical-Engineering's Last Exam) benchmark score

Browse files

Hi MiniMax team,

This PR adds `.eval_results/ple.yaml` with the [Physical-Engineering's Last Exam (PLE)](https://huggingface.co/datasets/google/physical-engineerings-last-exam) score for `MiniMaxAI/MiniMax-M2`.

- Score: `4.8` / 100 (std 1.8)
- Harness: Pi (the best of the 7 harnesses we ran with this model)
- Task id: `overall-score` (dev split)
- Results date: `2026-10-20`
- Leaderboard: https://ple-bench.com/leaderboard?set=dev
- Paper: https://arxiv.org/abs/2610.30537

PLE tests AI agents on structural, civil, electrical, and mechanical engineering tasks. The score is the mean reward over the 29 public dev tasks, averaged over 3 runs, with the model at its highest reasoning setting (where it has one) and a 6-hour limit per task. Each task runs in an isolated Harbor container with the project's files. A separate verifier grades the deliverables with automated checks, an expert rubric judged by Claude Opus 5.5, or both.

Thank you!

Files changed (1) hide show

.eval_results/ple.yaml +9 -0

.eval_results/ple.yaml ADDED Viewed

	@@ -0,0 +1,9 @@

+- dataset:
+    id: google/physical-engineerings-last-exam
+    task_id: overall-score
+  value: 4.8
+  date: "2026-10-20"
+  source:
+    url: https://ple-bench.com/leaderboard?set=dev
+    name: PLE leaderboard
+  notes: "Harness: Pi. Mean of 3 runs on the 29 dev tasks (std 1.8)."