Instructions to use tomtia/Intern-Decision-4B-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 tomtia/Intern-Decision-4B-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 tomtia/Intern-Decision-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf tomtia/Intern-Decision-4B-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf tomtia/Intern-Decision-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf tomtia/Intern-Decision-4B-GGUF:Q4_K_M

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 tomtia/Intern-Decision-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf tomtia/Intern-Decision-4B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/tomtia/Intern-Decision-4B-GGUF:Q4_K_M

Llama new
LM Studio
Jan

vLLM

How to use tomtia/Intern-Decision-4B-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tomtia/Intern-Decision-4B-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": "tomtia/Intern-Decision-4B-GGUF",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/tomtia/Intern-Decision-4B-GGUF:Q4_K_M

Ollama
How to use tomtia/Intern-Decision-4B-GGUF with Ollama:
```
ollama run hf.co/tomtia/Intern-Decision-4B-GGUF:Q4_K_M
```
Unsloth Desktop

How to use tomtia/Intern-Decision-4B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf tomtia/Intern-Decision-4B-GGUF:Q4_K_M

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": "tomtia/Intern-Decision-4B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use tomtia/Intern-Decision-4B-GGUF with Docker Model Runner:
```
docker model run hf.co/tomtia/Intern-Decision-4B-GGUF:Q4_K_M
```

Lemonade

How to use tomtia/Intern-Decision-4B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull tomtia/Intern-Decision-4B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Intern-Decision-4B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use tomtia/Intern-Decision-4B-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 tomtia/Intern-Decision-4B-GGUF:Q4_K_M

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 tomtia/Intern-Decision-4B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use tomtia/Intern-Decision-4B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf tomtia/Intern-Decision-4B-GGUF:Q4_K_M

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 "tomtia/Intern-Decision-4B-GGUF:Q4_K_M" \
  --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"

Intern-Decision-4B GGUF (Q4_K_M) for Tomtia

This repository holds the exact model pair that the Tomtia app downloads: a Q4_K_M quantization of internlm/Intern-Decision-4B and its unchanged vision projector. Tomtia is a local-first file organizer; the model only chooses among fixed options, on the user's own computer.

Files

File	Bytes	sha256
`intern-decision-4b-Q4_K_M.gguf`	2,708,804,512	`024340672f72795421c08cb23f74504cfebc77a5694367ae1d8d50698ebe92d0`
`mmproj-f16.gguf`	672,423,424	`223984a70cddd16ee7478212498ded608c32fb46e9aec31089e137bc95a6e9b7`

Use both files together. The app checks both hashes before it uses them.

Where the files come from

Qwen/Qwen3.5-4B (Apache-2.0, Alibaba Cloud).
internlm/Intern-Decision-4B, decision-tuned from Qwen3.5-4B (Apache-2.0, Shanghai AI Laboratory; the Qwen licence kept as LICENSE-QWEN).
bombdefuser-124/Intern-Decision-4B-GGUF at revision 0b1dc0b4a5b7eed290f662644237f5ee3694f260: GGUF conversion with corrected metadata (intern-decision-4b-fp16-fixed.gguf, sha256 8db59910ebba0329b5dd503b5a7747b398664388279627c1a9478f896479dfaf) and the vision projector mmproj-f16.gguf.
This repository: the FP16 file above, quantized by Tomtia to Q4_K_M. The projector is copied unchanged from step 3.

How the body was made (modification notice)

intern-decision-4b-Q4_K_M.gguf is a modified file: Tomtia quantized intern-decision-4b-fp16-fixed.gguf to Q4_K_M. No other change was made, and no importance matrix was used.

Tool: llama.cpp llama-quantize, version 0.6.0-dev, build 300, commit b42b7e6.
Command: llama-quantize intern-decision-4b-fp16-fixed.gguf intern-decision-4b-Q4_K_M.gguf Q4_K_M 4
First made on 2026-10-09. Rebuilt on 2026-11-03 from the public FP16 with the same build and command; the result was byte-identical (same sha256). The script is in rebuild.sh.

llama.cpp

llama-server \
  --model intern-decision-4b-Q4_K_M.gguf \
  --mmproj mmproj-f16.gguf

Licence

Apache License 2.0. See LICENSE (Intern-Decision, Shanghai AI Laboratory) and LICENSE-QWEN (Qwen, Alibaba Cloud). NOTICE lists the attributions and the change. When you redistribute these files, keep the licence files and the notices.

Versions

Every version stays as its own commit; history is never rewritten, because installed apps point at a fixed commit. A later trained version will be a new commit with its own hashes and notes.

Downloads last month: 146

GGUF

Model size

4B params

Architecture

qwen35

Hardware compatibility

4-bit

Model tree for tomtia/Intern-Decision-4B-GGUF

Base model

Qwen/Qwen3.5-4B-Base

Finetuned

Qwen/Qwen3.5-4B

Finetuned

internlm/Intern-Decision-4B

Quantized

(4)

this model