Instructions to use mlx-community/LongCat-Next-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mlx-community/LongCat-Next-4bit with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("mlx-community/LongCat-Next-4bit")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Transformers

How to use mlx-community/LongCat-Next-4bit with Transformers:

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

pipe = pipeline("text-generation", model="mlx-community/LongCat-Next-4bit", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("mlx-community/LongCat-Next-4bit", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

vLLM

How to use mlx-community/LongCat-Next-4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mlx-community/LongCat-Next-4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mlx-community/LongCat-Next-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mlx-community/LongCat-Next-4bit

SGLang

How to use mlx-community/LongCat-Next-4bit 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 "mlx-community/LongCat-Next-4bit" \
    --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": "mlx-community/LongCat-Next-4bit",
		"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 "mlx-community/LongCat-Next-4bit" \
        --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": "mlx-community/LongCat-Next-4bit",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

How to use mlx-community/LongCat-Next-4bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "mlx-community/LongCat-Next-4bit"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "mlx-community/LongCat-Next-4bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use mlx-community/LongCat-Next-4bit with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "mlx-community/LongCat-Next-4bit"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default mlx-community/LongCat-Next-4bit

Run Hermes

hermes

MLX LM

How to use mlx-community/LongCat-Next-4bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "mlx-community/LongCat-Next-4bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "mlx-community/LongCat-Next-4bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "mlx-community/LongCat-Next-4bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Docker Model Runner
How to use mlx-community/LongCat-Next-4bit with Docker Model Runner:
```
docker model run hf.co/mlx-community/LongCat-Next-4bit
```

A fun, quick little model!

by phakio - opened Mar 30

Discussion

phakio

Mar 30

Always like to see competition in the local LLM field. This model is quick, and due to it not being trained as a thinking model, produces a reply much quicker than the Qwen3.5 series. Sure you can disable thinking in the Qwen 3.5 series, but why not try out a different architecture!

Reading the release paper, this model has a little bit of everything, image analysis, text to speech, speech to speech, and this quant is only accessing the text generation side of it, but on my M1 Max Macbook Pro with 64gb ram, I've found that its generation is adequate for my needs, and quick enough to not feel sluggish.

Overall generations are without fluff and to the point. short, concise replies, but model will elaborate and explain things if needed. I like it.

((mlx_env) ) phone@phones-MacBook-Pro longcat-mlx % mlx_lm.generate \
    --model /Users/phone/.lmstudio/models/mlx-community/LongCat-Next-4bit \
    --prompt "Write a very long essay about probiscus monkeys." \
    --max-tokens 5000 \
    --trust-remote-code \
    --verbose VERBOSE
==========


--TRUNCATED TO NOT FILL THIS DISCUSSION WITH MONKEY FACTS--


==========
Prompt: 12 tokens, 2.615 tokens-per-sec
Generation: 2191 tokens, 59.284 tokens-per-sec
Peak memory: 38.674 GB
((mlx_env) ) phone@phones-MacBook-Pro longcat-mlx %

Prompt: 1046 tokens, 156.851 tokens-per-sec
Generation: 186 tokens, 59.222 tokens-per-sec
Peak memory: 39.457 GB
((mlx_env) ) phone@phones-MacBook-Pro longcat-mlx %

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment