Text Generation
MLX
Safetensors
English
Chinese
glm_moe_dsa
glm
glm-5
apple-silicon
quantized
2-8bit
Mixture of Experts
orcasaq
dynamic-quant
reasoning
coding
agentic
conversational
4-bit precision
Instructions to use orcarouter/GLM-5.3-MLX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use orcarouter/GLM-5.3-MLX 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("orcarouter/GLM-5.3-MLX") 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
- Pi
How to use orcarouter/GLM-5.3-MLX with Pi:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "orcarouter/GLM-5.3-MLX"
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "mlx-lm": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "orcarouter/GLM-5.3-MLX" } ] } } }Run Pi
# Start Pi in your project directory: pi
- MLX LM
How to use orcarouter/GLM-5.3-MLX with MLX LM:
Generate or start a chat session
# Install MLX LM uv tool install mlx-lm # Interactive chat REPL mlx_lm.chat --model "orcarouter/GLM-5.3-MLX"
Run an OpenAI-compatible server
# Install MLX LM uv tool install mlx-lm # Start the server mlx_lm.server --model "orcarouter/GLM-5.3-MLX" # Calling the OpenAI-compatible server with curl curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "orcarouter/GLM-5.3-MLX", "messages": [ {"role": "user", "content": "Hello"} ] }' - Hermes Agent
How to use orcarouter/GLM-5.3-MLX 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 "orcarouter/GLM-5.3-MLX"
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 orcarouter/GLM-5.3-MLX
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use orcarouter/GLM-5.3-MLX with OpenClaw:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "orcarouter/GLM-5.3-MLX"
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 "orcarouter/GLM-5.3-MLX" \ --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"
| { | |
| "attention@8bit": { | |
| "n": 390, | |
| "cos_mean": 0.9999843754304225, | |
| "cos_min": 0.9999701142781648, | |
| "cos_p5": 0.9999817950797752, | |
| "snr_mean": 45.05945505567317, | |
| "worst": [ | |
| "model.layers.0.self_attn.q_b_proj.weight", | |
| "model.layers.7.self_attn.q_b_proj.weight", | |
| "model.layers.2.self_attn.q_b_proj.weight", | |
| "model.layers.3.self_attn.q_b_proj.weight", | |
| "model.layers.3.self_attn.kv_a_proj_with_mqa.weight" | |
| ] | |
| }, | |
| "dense_mlp@6bit": { | |
| "n": 9, | |
| "cos_mean": 0.9997481794588633, | |
| "cos_min": 0.9997417464914627, | |
| "cos_p5": 0.9997417464914627, | |
| "snr_mean": 32.974673175379735, | |
| "worst": [ | |
| "model.layers.2.mlp.down_proj.weight", | |
| "model.layers.1.mlp.down_proj.weight", | |
| "model.layers.0.mlp.down_proj.weight", | |
| "model.layers.1.mlp.up_proj.weight", | |
| "model.layers.2.mlp.gate_proj.weight" | |
| ] | |
| }, | |
| "expert_down@5bit": { | |
| "n": 19200, | |
| "cos_mean": 0.9989847256501002, | |
| "cos_min": 0.9981335686868937, | |
| "cos_p5": 0.9989747346453054, | |
| "snr_mean": 26.926658932773485, | |
| "worst": [ | |
| "model.layers.4.mlp.experts.96.down_proj.weight", | |
| "model.layers.6.mlp.experts.87.down_proj.weight", | |
| "model.layers.3.mlp.experts.200.down_proj.weight", | |
| "model.layers.5.mlp.experts.234.down_proj.weight", | |
| "model.layers.8.mlp.experts.86.down_proj.weight" | |
| ] | |
| }, | |
| "expert_gate_up@4bit": { | |
| "n": 38400, | |
| "cos_mean": 0.9956952468956632, | |
| "cos_min": 0.9950926262273702, | |
| "cos_p5": 0.9955739241692824, | |
| "snr_mean": 20.655963162017926, | |
| "worst": [ | |
| "model.layers.4.mlp.experts.168.gate_proj.weight", | |
| "model.layers.3.mlp.experts.2.up_proj.weight", | |
| "model.layers.4.mlp.experts.168.up_proj.weight", | |
| "model.layers.3.mlp.experts.2.gate_proj.weight", | |
| "model.layers.4.mlp.experts.186.gate_proj.weight" | |
| ] | |
| }, | |
| "shared_expert@6bit": { | |
| "n": 225, | |
| "cos_mean": 0.9997443424916399, | |
| "cos_min": 0.9996723558344117, | |
| "cos_p5": 0.9997260646039723, | |
| "snr_mean": 32.91120458715047, | |
| "worst": [ | |
| "model.layers.77.mlp.shared_experts.down_proj.weight", | |
| "model.layers.74.mlp.shared_experts.down_proj.weight", | |
| "model.layers.73.mlp.shared_experts.down_proj.weight", | |
| "model.layers.76.mlp.shared_experts.down_proj.weight", | |
| "model.layers.75.mlp.shared_experts.down_proj.weight" | |
| ] | |
| } | |
| } |