Instructions to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-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 prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16
# Run inference directly in the terminal:
llama cli -hf prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16
# Run inference directly in the terminal:
llama cli -hf prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

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 prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16
# Run inference directly in the terminal:
./build/bin/llama cli -hf prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

Use Docker

docker model run hf.co/prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

Llama new
LM Studio
Jan

vLLM

How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-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": "prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

SGLang

How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with Ollama:
```
ollama run hf.co/prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16
```
Unsloth Desktop

How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

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": "prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with Docker Model Runner:
```
docker model run hf.co/prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16
```

Lemonade

How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

Run and chat with the model

lemonade run user.Clover-35B-A3B-Preview-MXFP4-MoE-GGUF-BF16

List all available models

lemonade list

Hermes Agent

How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-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 prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

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 prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16

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 "prithivMLmods/Clover-35B-A3B-Preview-MXFP4-MoE-GGUF:BF16" \
  --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"

Clover-35B-A3B-Preview-MXFP4-MoE-GGUF

Clover-35B-A3B-Preview is an open-source (Apache-2.0) block diffusion language model developed by DiffuSpace and derived from the Qwen3.6-35B-A3B architecture. Built on a 35B-parameter hybrid Gated DeltaNet (GDN) and full-attention Mixture-of-Experts (MoE) backbone that activates roughly 3B parameters per token across a 128K context window, this text-only model adapts traditional autoregressive generation into parallel block diffusion using bidirectional block attention, block-level expert coreset routing, and LowConfidence decoding (with additional support for self-speculative decoding via SGLang and llama.cpp). Equipped with native reasoning capabilities and built-in repetition and length penalties to ensure stable long-horizon outputs, Clover-35B-A3B-Preview achieves state-of-the-art results across agentic and coding benchmarks—including SWE-bench Verified (73.6), BFCL-v4 (67.9), Overall (65.3), Terminal-Bench 2.1 (51.7), SWE-Pro (46.5), and Terminal-Hard (38.1)—consistently outperforming both competing diffusion models (such as LLaDA2.2 flash and Mercury 2.5) and its own autoregressive Qwen3.5 and Qwen3.6 counterparts.

Model Files

File Name	Quant Type	File Size	File Link	Description
Clover-35B-A3B-Preview.BF16.gguf	BF16	69.4 GB	Link	Full BF16 weights. Highest quality, largest file size.
Clover-35B-A3B-Preview.MXFP4_MOE.gguf	MXFP4_MOE	19.8 GB	Link	MXFP4_MOE: Highly optimized for NVIDIA Blackwell hardware accelerators (e.g., RTX Pro 6000, RTX 50-series, B200, B300), offering native hardware-accelerated FP4 Tensor Core speeds and a strong balance of size and performance.
Clover-35B-A3B-Preview.Q3_K_L.gguf	Q3_K_L	18.1 GB	Link	Lower quality but usable, good for low RAM availability.
Clover-35B-A3B-Preview.Q5_K_M.gguf	Q5_K_M	24.7 GB	Link	High quality, recommended for standard hardware.
Clover-35B-A3B-Preview.Q6_K.gguf	Q6_K	28.5 GB	Link	Very high quality, near perfect, recommended for standard hardware.

MXFP4_MOE Test Run

Tested Clover-35B-A3B-Preview.MXFP4_MOE.gguf on an NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with 96 GB of VRAM, using llama-cli with full GPU offloading.

Performance Metrics

Metric	Value
Average Prompt Processing Speed	238.55 tokens/s
Average Generation Speed	167.81 tokens/s
Total Wall Time	107.8 seconds
Context Length	16,384 tokens
KV Cache (K)	`q8_0`
GPU	NVIDIA RTX PRO 6000 Blackwell Server Edition
GPU VRAM Capacity	96 GB
Peak GPU VRAM Usage	19.07 GiB
Peak GPU Utilization	100%
GPU Layers Offloaded	99 (`-ngl 99`)
Random Seed	42
Speculative Decoding	No
Test Runs	5

Test Shots

Shot	Test
Shot 1	C: Insert Node at End of Singly Linked List
Shot 2	Jogger and Train Passing Time
Shot 3	Two Pipes Filling a Tank
Shot 4	Students in Examination Rooms
Shot 5	Father and Son Ages

Run Inference

./build/bin/llama-cli \
  -m /path/to/models/Clover-35B-A3B-Preview.MXFP4_MOE.gguf \
  -ngl 99 \
  -c 16384 \
  --cache-type-k q8_0 \
  --seed 42 \
  -n 8192 \
  -st \
  -p "Your prompt here"

Performance results reflect this specific hardware and inference configuration. Actual throughput may vary depending on the prompt, model, software build, and runtime settings.