Instructions to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF",
	filename="Qwen3.5-0.8B/Qwen3.5-0.8B-PRISM-DQ.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16
# Run inference directly in the terminal:
llama-cli -hf Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16
# Run inference directly in the terminal:
llama-cli -hf Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

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 Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16
# Run inference directly in the terminal:
./llama-cli -hf Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

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 Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

Use Docker

docker model run hf.co/Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

LM Studio
Jan

vLLM

How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-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": "Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

Ollama
How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with Ollama:
```
ollama run hf.co/Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16
```

Unsloth Studio new

How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF 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 Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF 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 Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF to start chatting

Pi new

How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

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": "Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-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 Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

Run Hermes

hermes

Docker Model Runner
How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with Docker Model Runner:
```
docker model run hf.co/Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16
```

Lemonade

How to use Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF:BF16

Run and chat with the model

lemonade run user.Qwen3.5-PRISM-Dynamic-Quant-GGUF-BF16

List all available models

lemonade list

Qwen3.5 PRISM Dynamic Quantization (GGUF)

PRISM Dynamic Quantization (PRISM-DQ) applies per-tensor-class bit allocation based on structural weight analysis — no calibration data or importance matrices required. Each tensor class (attention keys, FFN gates, SSM components, etc.) receives a quantization type proportional to its measured sensitivity, while staying within a target bits-per-weight budget.

This repo contains PRISM-DQ quantized GGUFs for the full Qwen3.5 vision-language model family (0.8B, 2B, 4B, 9B), plus multimodal projection weights (mmproj) for vision capabilities.

Benchmark Results

Perplexity Comparison (UltraChat, 5 chunks, 512 ctx)

Model	Method	BPW	PPL	Size
Qwen3.5-0.8B	Q3_K_M	4.96	12.14	470 MB
	PRISM-DQ	4.94	11.42	468 MB
	Q3_K_M (imatrix)	4.96	11.31	470 MB
	UD-Q3_K_XL	5.19	10.94	492 MB
	IQ4_XS (imatrix)	5.20	10.35	493 MB
	UD-Q4_K_XL	5.89	10.07	559 MB
Qwen3.5-2B	Q3_K_M	4.69	9.35	1107 MB
	PRISM-DQ	4.68	9.26	1104 MB
	Q3_K_M (imatrix)	4.69	8.40	1107 MB
	UD-Q3_K_XL	4.91	8.27	1159 MB
	IQ4_XS (imatrix)	4.97	8.12	1173 MB
	UD-Q4_K_XL	5.68	8.07	1340 MB
Qwen3.5-4B	Q3_K_M	4.36	6.88	2293 MB
	PRISM-DQ	4.31	6.82	2271 MB
	Q3_K_M (imatrix)	4.36	6.62	2293 MB
	UD-Q3_K_XL	4.63	6.66	2436 MB
	IQ4_XS (imatrix)	4.70	6.51	2477 MB
	UD-Q4_K_XL	5.53	6.56	2912 MB
Qwen3.5-9B	Q3_K_M	4.17	6.25	4674 MB
	PRISM-DQ	4.15	6.18	4652 MB
	Q3_K_M (imatrix)	4.17	5.96	4674 MB
	UD-Q3_K_XL	4.51	6.01	5054 MB
	IQ4_XS (imatrix)	4.61	6.03	5169 MB
	UD-Q4_K_XL	5.33	5.86	5966 MB

Key Findings

PRISM-DQ beats uniform Q3_K_M on all 4 models (1-6% PPL improvement) at same or lower BPW
Smallest file size at competitive perplexity across the Qwen3.5 family
No calibration data needed — allocation decisions are purely weight-analysis-based
When combined with importance matrices, PRISM-DQ+imatrix achieves Pareto-optimal results on 4B and 9B

Model Files

Each subfolder contains the quantized model GGUF plus multimodal projection weights:

Qwen3.5-0.8B/
  Qwen3.5-0.8B-PRISM-DQ.gguf    (446 MB)
  mmproj-BF16.gguf
  mmproj-F16.gguf
  mmproj-F32.gguf
  chat_template.jinja

Qwen3.5-2B/
  Qwen3.5-2B-PRISM-DQ.gguf      (1.0 GB)
  mmproj-BF16.gguf
  mmproj-F16.gguf
  mmproj-F32.gguf
  chat_template.jinja

Qwen3.5-4B/
  Qwen3.5-4B-PRISM-DQ.gguf      (2.1 GB)
  mmproj-BF16.gguf
  mmproj-F16.gguf
  mmproj-F32.gguf
  chat_template.jinja

Qwen3.5-9B/
  Qwen3.5-9B-PRISM-DQ.gguf      (4.3 GB)
  mmproj-BF16.gguf
  mmproj-F16.gguf
  mmproj-F32.gguf
  chat_template.jinja

Usage

Text-only (llama.cpp)

# Download a model
huggingface-cli download Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF \
  Qwen3.5-9B/Qwen3.5-9B-PRISM-DQ.gguf --local-dir .

# Run with llama-cli
llama-cli -m Qwen3.5-9B/Qwen3.5-9B-PRISM-DQ.gguf \
  -p "You are a helpful assistant." \
  --chat-template-file Qwen3.5-9B/chat_template.jinja \
  -cnv

Vision (multimodal)

# Download model + mmproj
huggingface-cli download Ex0bit/Qwen3.5-PRISM-Dynamic-Quant-GGUF \
  Qwen3.5-9B/Qwen3.5-9B-PRISM-DQ.gguf \
  Qwen3.5-9B/mmproj-BF16.gguf --local-dir .

# Run with llama-mtmd-cli
llama-mtmd-cli -m Qwen3.5-9B/Qwen3.5-9B-PRISM-DQ.gguf \
  --mmproj Qwen3.5-9B/mmproj-BF16.gguf \
  --chat-template-file Qwen3.5-9B/chat_template.jinja \
  -cnv

LM Studio / Ollama

These GGUFs work with any llama.cpp-compatible runtime. Simply point your application at the .gguf file.

PRISM-DQ Quantization Recipes

Qwen3.5-0.8B (target 3.5 BPW)

llama-quantize \
  --tensor-type "attn_gate=Q3_K" \
  --tensor-type "attn_k=Q3_K" \
  --tensor-type "attn_output=IQ4_XS" \
  --tensor-type "attn_q=Q3_K" \
  --tensor-type "attn_qkv=Q3_K" \
  --tensor-type "attn_v=Q4_K" \
  --tensor-type "ffn_down=Q3_K" \
  --tensor-type "ffn_gate=Q3_K" \
  --tensor-type "ffn_up=Q3_K" \
  --tensor-type "ssm_alpha=Q3_K" \
  --tensor-type "ssm_beta=IQ4_XS" \
  --tensor-type "ssm_out=IQ4_XS" \
  --tensor-type "token_embd=Q3_K" \
  --tensor-type "blk\.(4)\.ssm_beta=Q4_K" \
  --tensor-type "blk\.(18)\.ssm_out=Q4_K" \
  input.gguf output.gguf Q3_K

Qwen3.5-2B (target 3.5 BPW)

llama-quantize \
  --tensor-type "attn_gate=Q3_K" \
  --tensor-type "attn_k=Q4_K" \
  --tensor-type "attn_output=Q4_K" \
  --tensor-type "attn_q=Q4_K" \
  --tensor-type "attn_qkv=Q3_K" \
  --tensor-type "attn_v=Q4_K" \
  --tensor-type "ffn_down=Q3_K" \
  --tensor-type "ffn_gate=Q3_K" \
  --tensor-type "ffn_up=Q3_K" \
  --tensor-type "ssm_alpha=Q4_K" \
  --tensor-type "ssm_beta=Q4_K" \
  --tensor-type "ssm_out=Q3_K" \
  --tensor-type "token_embd=Q3_K" \
  input.gguf output.gguf Q3_K

Qwen3.5-4B (target 3.5 BPW)

llama-quantize \
  --tensor-type "attn_gate=Q3_K" \
  --tensor-type "attn_k=Q4_K" \
  --tensor-type "attn_output=Q5_K" \
  --tensor-type "attn_q=Q3_K" \
  --tensor-type "attn_qkv=Q3_K" \
  --tensor-type "attn_v=Q4_K" \
  --tensor-type "ffn_down=Q3_K" \
  --tensor-type "ffn_gate=Q3_K" \
  --tensor-type "ffn_up=Q3_K" \
  --tensor-type "ssm_alpha=Q4_K" \
  --tensor-type "ssm_beta=Q4_K" \
  --tensor-type "ssm_out=Q3_K" \
  --tensor-type "token_embd=Q3_K" \
  input.gguf output.gguf Q3_K

Qwen3.5-9B (target 3.5 BPW)

llama-quantize \
  --tensor-type "attn_gate=Q3_K" \
  --tensor-type "attn_k=Q4_K" \
  --tensor-type "attn_output=IQ4_XS" \
  --tensor-type "attn_q=Q4_K" \
  --tensor-type "attn_qkv=Q3_K" \
  --tensor-type "attn_v=Q4_K" \
  --tensor-type "ffn_down=Q3_K" \
  --tensor-type "ffn_gate=Q3_K" \
  --tensor-type "ffn_up=Q3_K" \
  --tensor-type "output=Q3_K" \
  --tensor-type "ssm_alpha=Q4_K" \
  --tensor-type "ssm_beta=Q4_K" \
  --tensor-type "ssm_out=Q3_K" \
  --tensor-type "token_embd=Q3_K" \
  input.gguf output.gguf Q3_K

How PRISM-DQ Works

PRISM Dynamic Quantization analyzes each weight tensor using 7 structural metrics:

PL-Alpha-Hill — spectral heavy-tail index via eigenvalue analysis
Spectral Dominance — top singular value ratio (rank-1 approximation quality)
OSQE — optimal scale quantization error at multiple bit levels (2, 3, 4, 6 bit)
Matrix Imbalance — max of row/column coefficient of variation
Fragility — log-ratio of 2-bit vs 4-bit quantization error
Boundary Density — fraction of values near quantization bin boundaries
Spectral Position Prior — bidirectional spectral norm product encoding layer position

These metrics are combined into a composite sensitivity score per tensor class. A Lagrangian allocator then distributes bits across classes to minimize total quantization distortion subject to the BPW budget, with per-block refinement for individual tensor overrides.