Instructions to use spicyneuron/Kimi-K2.7-Code-MLX-3.6bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use spicyneuron/Kimi-K2.7-Code-MLX-3.6bit 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("spicyneuron/Kimi-K2.7-Code-MLX-3.6bit")

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 spicyneuron/Kimi-K2.7-Code-MLX-3.6bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "spicyneuron/Kimi-K2.7-Code-MLX-3.6bit"

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": "spicyneuron/Kimi-K2.7-Code-MLX-3.6bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use spicyneuron/Kimi-K2.7-Code-MLX-3.6bit 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 "spicyneuron/Kimi-K2.7-Code-MLX-3.6bit"

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 spicyneuron/Kimi-K2.7-Code-MLX-3.6bit

Run Hermes

hermes

MLX LM

How to use spicyneuron/Kimi-K2.7-Code-MLX-3.6bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "spicyneuron/Kimi-K2.7-Code-MLX-3.6bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "spicyneuron/Kimi-K2.7-Code-MLX-3.6bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "spicyneuron/Kimi-K2.7-Code-MLX-3.6bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Kimi-K2.7-Code-MLX-3.6bit / README.md

spicyneuron

Update README.md

f2f75bb verified 13 days ago

preview code

Raw

History Blame Contribute Delete

1.3 kB

	---
	language: en
	pipeline_tag: text-generation
	library_name: mlx
	base_model:
	- moonshotai/Kimi-K2.7-Code
	base_model_relation: quantized
	tags:
	- mlx
	---

	[moonshotai/Kimi-K2.7-Code](https://huggingface.co/moonshotai/Kimi-K2.7-Code)
	optimized for running on a Mac Studio M3 Ultra.

	- A mixed-precision quant that balances speed, memory, and accuracy.
	- 3-bit MoE baseline with important always-on layers at higher precision.
	- Fits into ~460 GB memory, leaving enough room for a smaller utility model.

	# Usage

	```sh
	# Start server at http://localhost:8080/v1/chat/completions
	uvx --from mlx-lm mlx_lm.server \
	--host 127.0.0.1 \
	--port 8080 \
	--model spicyneuron/Kimi-K2.7-Code-MLX-3.6bit
	```

	# Benchmarks

	metric \| this model
	--- \| ---
	bpw \| 3.578
	base memory \| 427.579
	peak memory (1024/512) \| 460.444
	prompt tok/s (1024) \| 218.851 ± 0.208
	gen tok/s (512) \| 21.035 ± 0.049
	perplexity \| 4.462 ± 0.037
	arc_challenge \| 0.692 ± 0.021
	hellaswag \| 0.780 ± 0.019

	# Methodology

	Quantized with a [mlx-lm fork](https://github.com/spicyneuron/mlx-lm/tree/_tools).
	MLX quantization options differ than llama.cpp, but the principles are the same:

	- Sensitive layers like MoE routing, attention, and output embeddings get higher precision
	- More tolerant layers like MoE experts get lower precision