Instructions to use anikifoss/GLM-4.5-Air-HQ4_K with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use anikifoss/GLM-4.5-Air-HQ4_K with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf anikifoss/GLM-4.5-Air-HQ4_K
# Run inference directly in the terminal:
llama cli -hf anikifoss/GLM-4.5-Air-HQ4_K

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf anikifoss/GLM-4.5-Air-HQ4_K
# Run inference directly in the terminal:
llama cli -hf anikifoss/GLM-4.5-Air-HQ4_K

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf anikifoss/GLM-4.5-Air-HQ4_K
# Run inference directly in the terminal:
./llama-cli -hf anikifoss/GLM-4.5-Air-HQ4_K

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf anikifoss/GLM-4.5-Air-HQ4_K
# Run inference directly in the terminal:
./build/bin/llama-cli -hf anikifoss/GLM-4.5-Air-HQ4_K

Use Docker

docker model run hf.co/anikifoss/GLM-4.5-Air-HQ4_K

LM Studio
Jan

vLLM

How to use anikifoss/GLM-4.5-Air-HQ4_K with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "anikifoss/GLM-4.5-Air-HQ4_K"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "anikifoss/GLM-4.5-Air-HQ4_K",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/anikifoss/GLM-4.5-Air-HQ4_K

Ollama
How to use anikifoss/GLM-4.5-Air-HQ4_K with Ollama:
```
ollama run hf.co/anikifoss/GLM-4.5-Air-HQ4_K
```

Unsloth Studio

How to use anikifoss/GLM-4.5-Air-HQ4_K with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for anikifoss/GLM-4.5-Air-HQ4_K to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for anikifoss/GLM-4.5-Air-HQ4_K to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for anikifoss/GLM-4.5-Air-HQ4_K to start chatting

How to use anikifoss/GLM-4.5-Air-HQ4_K with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf anikifoss/GLM-4.5-Air-HQ4_K

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "anikifoss/GLM-4.5-Air-HQ4_K"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

OpenClaw new

How to use anikifoss/GLM-4.5-Air-HQ4_K with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf anikifoss/GLM-4.5-Air-HQ4_K

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 "anikifoss/GLM-4.5-Air-HQ4_K" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

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

Docker Model Runner
How to use anikifoss/GLM-4.5-Air-HQ4_K with Docker Model Runner:
```
docker model run hf.co/anikifoss/GLM-4.5-Air-HQ4_K
```

Lemonade

How to use anikifoss/GLM-4.5-Air-HQ4_K with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull anikifoss/GLM-4.5-Air-HQ4_K

Run and chat with the model

lemonade run user.GLM-4.5-Air-HQ4_K-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use anikifoss/GLM-4.5-Air-HQ4_K with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf anikifoss/GLM-4.5-Air-HQ4_K

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 anikifoss/GLM-4.5-Air-HQ4_K

Run Hermes

hermes

Atomic Chat

anikifoss commited on Aug 6, 2025

Commit

38c2ba8

verified ·

1 Parent(s): c17ddd0

Update README.md

Browse files

Files changed (1) hide show

README.md +13 -13

README.md CHANGED Viewed

@@ -67,7 +67,7 @@ custom="
 blk\.0\.ffn_down\.weight=q8_0
 blk\.0\.ffn_gate\.weight=q8_0
 blk\.0\.ffn_up\.weight=q8_0
-blk\.46\.nextn\.eh_proj\.weight=f16
 blk\.46\.nextn\.embed_tokens\.weight=q8_0
 blk\.46\.nextn\.enorm\.weight=f32
 blk\.46\.nextn\.hnorm\.weight=f32
@@ -104,32 +104,32 @@ blk\.4[0-6]\.post_attention_norm\.weight=f32
 blk\.[1-9]\.exp_probs_b\.bias=f32
 blk\.[1-9]\.ffn_down_exps\.weight=q8_0
-blk\.[1-9]\.ffn_down_shexp\.weight=f16
 blk\.[1-9]\.ffn_gate_exps\.weight=q4_K
 blk\.[1-9]\.ffn_gate_inp\.weight=f32
-blk\.[1-9]\.ffn_gate_shexp\.weight=f16
 blk\.[1-9]\.ffn_up_exps\.weight=q4_K
-blk\.[1-9]\.ffn_up_shexp\.weight=f16
 blk\.[1-3][0-9]\.exp_probs_b\.bias=f32
 blk\.[1-3][0-9]\.ffn_down_exps\.weight=q8_0
-blk\.[1-3][0-9]\.ffn_down_shexp\.weight=f16
 blk\.[1-3][0-9]\.ffn_gate_exps\.weight=q4_K
 blk\.[1-3][0-9]\.ffn_gate_inp\.weight=f32
-blk\.[1-3][0-9]\.ffn_gate_shexp\.weight=f16
 blk\.[1-3][0-9]\.ffn_up_exps\.weight=q4_K
-blk\.[1-3][0-9]\.ffn_up_shexp\.weight=f16
 blk\.4[0-6]\.exp_probs_b\.bias=f32
 blk\.4[0-6]\.ffn_down_exps\.weight=q8_0
-blk\.4[0-6]\.ffn_down_shexp\.weight=f16
 blk\.4[0-6]\.ffn_gate_exps\.weight=q4_K
 blk\.4[0-6]\.ffn_gate_inp\.weight=f32
-blk\.4[0-6]\.ffn_gate_shexp\.weight=f16
 blk\.4[0-6]\.ffn_up_exps\.weight=q4_K
-blk\.4[0-6]\.ffn_up_shexp\.weight=f16
-output\.weight=f16
 output_norm\.weight=f32
-token_embd\.weight=f16
 "
 custom=$(
@@ -143,7 +143,7 @@ TARGET_MODEL="GLM-4.5-Air-HQ4_K"
 mkdir -p ~/Env/models/anikifoss/$TARGET_MODEL
 ./build/bin/llama-quantize \
     --custom-q "$custom" \
-    /mnt/data/Models/zai-org/GLM-4.5-Air-GGUF/GLM-4.5-Air-128x9.4B-F16-00001-of-00005.gguf \
     ~/Env/models/anikifoss/$TARGET_MODEL/$TARGET_MODEL.gguf \
     Q4_K \
     32

 blk\.0\.ffn_down\.weight=q8_0
 blk\.0\.ffn_gate\.weight=q8_0
 blk\.0\.ffn_up\.weight=q8_0
+blk\.46\.nextn\.eh_proj\.weight=bf16
 blk\.46\.nextn\.embed_tokens\.weight=q8_0
 blk\.46\.nextn\.enorm\.weight=f32
 blk\.46\.nextn\.hnorm\.weight=f32
 blk\.[1-9]\.exp_probs_b\.bias=f32
 blk\.[1-9]\.ffn_down_exps\.weight=q8_0
+blk\.[1-9]\.ffn_down_shexp\.weight=bf16
 blk\.[1-9]\.ffn_gate_exps\.weight=q4_K
 blk\.[1-9]\.ffn_gate_inp\.weight=f32
+blk\.[1-9]\.ffn_gate_shexp\.weight=bf16
 blk\.[1-9]\.ffn_up_exps\.weight=q4_K
+blk\.[1-9]\.ffn_up_shexp\.weight=bf16
 blk\.[1-3][0-9]\.exp_probs_b\.bias=f32
 blk\.[1-3][0-9]\.ffn_down_exps\.weight=q8_0
+blk\.[1-3][0-9]\.ffn_down_shexp\.weight=bf16
 blk\.[1-3][0-9]\.ffn_gate_exps\.weight=q4_K
 blk\.[1-3][0-9]\.ffn_gate_inp\.weight=f32
+blk\.[1-3][0-9]\.ffn_gate_shexp\.weight=bf16
 blk\.[1-3][0-9]\.ffn_up_exps\.weight=q4_K
+blk\.[1-3][0-9]\.ffn_up_shexp\.weight=bf16
 blk\.4[0-6]\.exp_probs_b\.bias=f32
 blk\.4[0-6]\.ffn_down_exps\.weight=q8_0
+blk\.4[0-6]\.ffn_down_shexp\.weight=bf16
 blk\.4[0-6]\.ffn_gate_exps\.weight=q4_K
 blk\.4[0-6]\.ffn_gate_inp\.weight=f32
+blk\.4[0-6]\.ffn_gate_shexp\.weight=bf16
 blk\.4[0-6]\.ffn_up_exps\.weight=q4_K
+blk\.4[0-6]\.ffn_up_shexp\.weight=bf16
+output\.weight=bf16
 output_norm\.weight=f32
+token_embd\.weight=bf16
 "
 custom=$(
 mkdir -p ~/Env/models/anikifoss/$TARGET_MODEL
 ./build/bin/llama-quantize \
     --custom-q "$custom" \
+    /mnt/data/Models/zai-org/GLM-4.5-Air-GGUF/GLM-4.5-Air-128x9.4B-BF16-00001-of-00005.gguf \
     ~/Env/models/anikifoss/$TARGET_MODEL/$TARGET_MODEL.gguf \
     Q4_K \
     32