Instructions to use MC7ever/MiniCPM5-1B-Agent-mlx-q2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MC7ever/MiniCPM5-1B-Agent-mlx-q2 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("MC7ever/MiniCPM5-1B-Agent-mlx-q2")

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)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use MC7ever/MiniCPM5-1B-Agent-mlx-q2 with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "MC7ever/MiniCPM5-1B-Agent-mlx-q2"

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": "MC7ever/MiniCPM5-1B-Agent-mlx-q2"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use MC7ever/MiniCPM5-1B-Agent-mlx-q2 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 "MC7ever/MiniCPM5-1B-Agent-mlx-q2"

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 MC7ever/MiniCPM5-1B-Agent-mlx-q2

Run Hermes

hermes

OpenClaw new

How to use MC7ever/MiniCPM5-1B-Agent-mlx-q2 with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "MC7ever/MiniCPM5-1B-Agent-mlx-q2"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "MC7ever/MiniCPM5-1B-Agent-mlx-q2" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

MLX LM

How to use MC7ever/MiniCPM5-1B-Agent-mlx-q2 with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "MC7ever/MiniCPM5-1B-Agent-mlx-q2"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "MC7ever/MiniCPM5-1B-Agent-mlx-q2"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "MC7ever/MiniCPM5-1B-Agent-mlx-q2",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

MiniCPM5-1B-Agent (MLX 2-bit)

Aggressively quantized MLX version of Luminia/MiniCPM5-1B-Agent-GGUF. Quantized with mlx_lm.convert using affine mode (group_size=32, 3.0 bits/weight). Smallest variant — ~387 MB, runs on constrained Apple Silicon.

Note: 2-bit quantization significantly impacts output quality. Use the Q4 variant for better results if memory allows.

About the model

MiniCPM5-1B-Agent is a tiny agentic coding agent for CPU: a full fine-tune of openbmb/MiniCPM5-1B specialized to reason in <think>, call a small tool set (bash/read/write/edit/glob/grep), and run → read output → debug → patch → verify.

Base model: openbmb/MiniCPM5-1B (RL+OPD checkpoint)
Architecture: LlamaForCausalLM — 24 layers, 16 attention heads (GQA), 1536 hidden, 130560 vocab
Parameters: 1,080,632,832 (quantized to ~3.0 bits/weight)
Quantization: affine, group_size=32, 2 bits
License: Apache-2.0

How to use

from mlx_lm import load, generate

model, tokenizer = load("MC7ever/MiniCPM5-1B-Agent-mlx-q2")

messages = [
    {"role": "user", "content": "Write a Python function to check if a number is prime."}
]
prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
response = generate(model, tokenizer, prompt=prompt, max_tokens=1024)
print(response)

Credits

Original model: Luminia/MiniCPM5-1B-Agent-GGUF by Luminia
Base model: openbmb/MiniCPM5-1B by OpenBMB — MiniCPM4 paper (arXiv:2506.07900)
Quantization: MLX + mlx-lm

Other variants

Variant	Size	Bits/Weight	Repo
Safetensors (fp16)	2.0 GB	16	MC7ever/MiniCPM5-1B-Agent-safetensors
MLX Q4	580 MB	4.5	MC7ever/MiniCPM5-1B-Agent-mlx-q4
MLX Q2	387 MB	3.0	This repo

Downloads last month: 23

Safetensors

Model size

0.1B params

Tensor type

F16

U32

MLX

Hardware compatibility

2-bit

Model tree for MC7ever/MiniCPM5-1B-Agent-mlx-q2

Base model

openbmb/MiniCPM5-1B

Quantized

Luminia/MiniCPM5-1B-Agent-GGUF

Quantized

(2)

this model

Paper for MC7ever/MiniCPM5-1B-Agent-mlx-q2

MiniCPM4: Ultra-Efficient LLMs on End Devices

Paper • 2506.07900 • Published Jun 9, 2025 • 101