Instructions to use mudler/Qwen3.5-35B-A3B-APEX-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use mudler/Qwen3.5-35B-A3B-APEX-GGUF 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 mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16
# Run inference directly in the terminal:
./build/bin/llama cli -hf mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

Use Docker

docker model run hf.co/mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

Llama new
LM Studio
Jan

vLLM

How to use mudler/Qwen3.5-35B-A3B-APEX-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mudler/Qwen3.5-35B-A3B-APEX-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mudler/Qwen3.5-35B-A3B-APEX-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

Ollama
How to use mudler/Qwen3.5-35B-A3B-APEX-GGUF with Ollama:
```
ollama run hf.co/mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16
```
Unsloth Desktop

How to use mudler/Qwen3.5-35B-A3B-APEX-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use mudler/Qwen3.5-35B-A3B-APEX-GGUF with Docker Model Runner:
```
docker model run hf.co/mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16
```

Lemonade

How to use mudler/Qwen3.5-35B-A3B-APEX-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

Run and chat with the model

lemonade run user.Qwen3.5-35B-A3B-APEX-GGUF-F16

List all available models

lemonade list

Hermes Agent

How to use mudler/Qwen3.5-35B-A3B-APEX-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

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 mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use mudler/Qwen3.5-35B-A3B-APEX-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16

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 "mudler/Qwen3.5-35B-A3B-APEX-GGUF:F16" \
  --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"

mudler commited on Apr 1

Commit

36aea14

verified ·

1 Parent(s): 3b84bdb

Upload README.md with huggingface_hub

Browse files

Files changed (1) hide show

README.md +8 -3

README.md CHANGED Viewed

@@ -17,13 +17,17 @@ quantized_by: apex-quant
 pipeline_tag: text-generation
 ---
-# Qwen3.5-35B-A3B APEX GGUF -- MoE-Aware Mixed-Precision Quantization
 **Brought to you by the [LocalAI](https://github.com/mudler/LocalAI) team** -- the creators of the free, open-source OpenAI-compatible API for running AI locally.
-MoE-aware mixed-precision GGUF quantizations of [Qwen3.5-35B-A3B](https://huggingface.co/Qwen/Qwen3.5-35B-A3B) using llama.cpp. APEX (Adaptive Precision for EXpert Models) assigns quantization precision per tensor type and per layer, exploiting the structural sparsity of Mixture-of-Experts models to achieve lossless compression that uniform quantization cannot.
-This repository contains seven APEX GGUF files covering every deployment scenario from maximum accuracy to consumer GPU inference. The best configuration (APEX Quality) **beats both Q8_0 and F16 perplexity while being 38% smaller than Q8_0**. I-variants use a diverse imatrix (chat, code, reasoning, tool-calling -- no Wikipedia) that trades tiny perplexity increases for significant accuracy gains and lower KL divergence.
 ## Available Files
@@ -36,6 +40,7 @@ This repository contains seven APEX GGUF files covering every deployment scenari
 | `Qwen3.5-35B-A3B-APEX-Compact.gguf` | **APEX Compact** | 16.1 GB | 6.783 | 69.8 t/s | Consumer 24 GB GPUs |
 | `Qwen3.5-35B-A3B-APEX-I-Compact.gguf` | **APEX I-Compact** | 16.1 GB | 6.669 | 69.8 t/s | 16 GB GPUs, best accuracy at this size |
 | `Qwen3.5-35B-A3B-APEX-Mini.gguf` | **APEX Mini** | 12.2 GB | 7.088 | **74.4 t/s** | Consumer 16 GB VRAM, smallest viable |
 **APEX Quality** uses a 3-tier layer-wise precision gradient (Q6_K/Q5_K/IQ4_XS) with Q8_0 shared experts. It achieves the lowest perplexity of any quantization tested -- beating even F16 (6.527 vs 6.537).

 pipeline_tag: text-generation
 ---
+# Qwen3.5-35B-A3B APEX GGUF -- A Novel MoE-Aware Mixed-Precision Quantization Technique
 **Brought to you by the [LocalAI](https://github.com/mudler/LocalAI) team** -- the creators of the free, open-source OpenAI-compatible API for running AI locally.
+**[APEX Technical Report](https://github.com/mudler/apex-quant/blob/main/paper/APEX_Technical_Report.md)** | **[GitHub Repository](https://github.com/mudler/apex-quant)** | **[LocalAI](https://github.com/mudler/LocalAI)**
+APEX (Adaptive Precision for EXpert Models) is a **novel quantization technique** for Mixture-of-Experts language models. Unlike uniform quantization methods that apply the same precision to every tensor, APEX introduces a **layer-wise precision gradient** combined with **MoE-aware tensor classification** and **diverse imatrix calibration** to achieve Q8_0-level quality at a fraction of the size. The method was discovered through systematic autonomous experimentation across 25+ quantization strategies.
+This repository contains seven APEX GGUF files plus a vision projector (mmproj) covering every deployment scenario from maximum accuracy to consumer GPU inference. The best configuration (APEX Quality) **beats both Q8_0 and F16 perplexity while being 38% smaller than Q8_0**. I-variants use a diverse imatrix (chat, code, reasoning, tool-calling -- no Wikipedia) that trades tiny perplexity increases for significant accuracy gains and lower KL divergence.
+For the full technical details, method description, and reproduction scripts, see the **[APEX GitHub repository](https://github.com/mudler/apex-quant)**.
 ## Available Files
 | `Qwen3.5-35B-A3B-APEX-Compact.gguf` | **APEX Compact** | 16.1 GB | 6.783 | 69.8 t/s | Consumer 24 GB GPUs |
 | `Qwen3.5-35B-A3B-APEX-I-Compact.gguf` | **APEX I-Compact** | 16.1 GB | 6.669 | 69.8 t/s | 16 GB GPUs, best accuracy at this size |
 | `Qwen3.5-35B-A3B-APEX-Mini.gguf` | **APEX Mini** | 12.2 GB | 7.088 | **74.4 t/s** | Consumer 16 GB VRAM, smallest viable |
+| `mmproj-F16.gguf` | **Vision Projector** | 899 MB | -- | -- | Required for vision/multimodal tasks |
 **APEX Quality** uses a 3-tier layer-wise precision gradient (Q6_K/Q5_K/IQ4_XS) with Q8_0 shared experts. It achieves the lowest perplexity of any quantization tested -- beating even F16 (6.527 vs 6.537).