Instructions to use mindlab-research/Macaron-V1-Preview-749B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mindlab-research/Macaron-V1-Preview-749B with Transformers:

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

pipe = pipeline("text-generation", model="mindlab-research/Macaron-V1-Preview-749B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mindlab-research/Macaron-V1-Preview-749B")
model = AutoModelForCausalLM.from_pretrained("mindlab-research/Macaron-V1-Preview-749B")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use mindlab-research/Macaron-V1-Preview-749B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mindlab-research/Macaron-V1-Preview-749B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mindlab-research/Macaron-V1-Preview-749B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mindlab-research/Macaron-V1-Preview-749B

SGLang

How to use mindlab-research/Macaron-V1-Preview-749B 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 "mindlab-research/Macaron-V1-Preview-749B" \
    --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": "mindlab-research/Macaron-V1-Preview-749B",
		"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 "mindlab-research/Macaron-V1-Preview-749B" \
        --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": "mindlab-research/Macaron-V1-Preview-749B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use mindlab-research/Macaron-V1-Preview-749B with Docker Model Runner:
```
docker model run hf.co/mindlab-research/Macaron-V1-Preview-749B
```

usermma commited on 21 days ago

Commit

3d4d564

verified ·

1 Parent(s): 33a6452

Update README.md

Browse files

parsing the text into better one..

Files changed (1) hide show

README.md +14 -14

README.md CHANGED Viewed

@@ -137,20 +137,20 @@ The headline benchmark suite focuses on personal-agent behavior, daily-life task
 | Category | Benchmark | Macaron V1 Preview | GLM 5.1 | GPT 5.4 | Claude Opus 4.6 | Gemini 3.1 Pro | Qwen 3.6 Plus | Minimax 2.7 |
 | --- | --- | --- | --- | --- | --- | --- | --- | --- |
-| Personal Agent Benchmark | Macaron Livingbench | 75.2 | 63.2 | 66.5 | 68.9 | 57.6 | 59.0 | 58.2 |
-|  | VitaBench | 59.6 | 56.8 | 48.7 | 53.0 | 55.2 | 47.5 | 52.2 |
-|  | VitaBench (Delivery) | 67.0 | 64.2 | 50.0 | 65.0 | 63.0 | 58.0 | 63.0 |
-|  | VitaBench (In-Store) | 75.0 | 70.0 | 55.0 | 66.0 | 68.0 | 58.0 | 58.0 |
-|  | VitaBench (OTA) | 51.0 | 54.0 | 41.0 | 45.0 | 48.0 | 36.0 | 51.0 |
-|  | VitaBench (Cross-shop) | 45.3 | 39.0 | -- | 36.0 | 42.0 | 38.0 | 37.0 |
-|  | A2UI-Bench | 75.6 | 61.7 | 74.1 | 67.6 | 71.0 | 69.8 | 54.4 |
-|  | A2UI L1 | 89.5 | 72.2 | 82.3 | 81.5 | 85.1 | 84.1 | 75.1 |
-|  | A2UI L2 | 67.2 | 54.7 | 71.8 | 59.4 | 64.1 | 59.9 | 46.3 |
-|  | A2UI L3 | 65.7 | 54.5 | 65.4 | 57.5 | 59.2 | 60.7 | 34.8 |
-|  | PinchBench | 92.5 | 76.6 | 88.4 | 88.9 | 82.9 | 85.9 | 84.5 |
-| General Agent Benchmark | Tau3 Bench | 67.6 | 70.6 | 72.9 | 72.4 | 67.1 | 70.7 | 67.6 |
-|  | SWE-bench Verified | 78.1 | 76.4 | 78.2 | 78.2 | 78.8 | 73.4 | 73.8 |
-|  | Terminal-Bench 2.0 | 67.4 | 63.5 | 75.1 | 65.4 | 68.5 | 61.6 | 57.0 |
 Higher is better for all scores shown in the charts and table.

 | Category | Benchmark | Macaron V1 Preview | GLM 5.1 | GPT 5.4 | Claude Opus 4.6 | Gemini 3.1 Pro | Qwen 3.6 Plus | Minimax 2.7 |
 | --- | --- | --- | --- | --- | --- | --- | --- | --- |
+| Personal Agent Benchmark | Macaron Livingbench | **75.2** | 63.2 | 66.5 | 68.9 | 57.6 | 59.0 | 58.2 |
+|  | VitaBench | **59.6** | 56.8 | 48.7 | 53.0 | 55.2 | 47.5 | 52.2 |
+|  | VitaBench (Delivery) | **67.0** | 64.2 | 50.0 | 65.0 | 63.0 | 58.0 | 63.0 |
+|  | VitaBench (In-Store) | **75.0** | 70.0 | 55.0 | 66.0 | 68.0 | 58.0 | 58.0 |
+|  | VitaBench (OTA) | 51.0 | **54.0** | 41.0 | 45.0 | 48.0 | 36.0 | 51.0 |
+|  | VitaBench (Cross-shop) | **45.3** | 39.0 | -- | 36.0 | 42.0 | 38.0 | 37.0 |
+|  | A2UI-Bench | **75.6** | 61.7 | 74.1 | 67.6 | 71.0 | 69.8 | 54.4 |
+|  | A2UI L1 | **89.5** | 72.2 | 82.3 | 81.5 | 85.1 | 84.1 | 75.1 |
+|  | A2UI L2 | 67.2 | 54.7 | **71.8** | 59.4 | 64.1 | 59.9 | 46.3 |
+|  | A2UI L3 | **65.7** | 54.5 | 65.4 | 57.5 | 59.2 | 60.7 | 34.8 |
+|  | PinchBench | **92.5** | 76.6 | 88.4 | 88.9 | 82.9 | 85.9 | 84.5 |
+| General Agent Benchmark | Tau3 Bench | 67.6 | 70.6 | **72.9** | 72.4 | 67.1 | 70.7 | 67.6 |
+|  | SWE-bench Verified | 78.1 | 76.4 | 78.2 | 78.2 | **78.8** | 73.4 | 73.8 |
+|  | Terminal-Bench 2.0 | 67.4 | 63.5 | **75.1** | 65.4 | 68.5 | 61.6 | 57.0 |
 Higher is better for all scores shown in the charts and table.