Instructions to use ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-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 ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-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 ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M
# Run inference directly in the terminal:
llama cli -hf ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M
# Run inference directly in the terminal:
llama cli -hf ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M

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 ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M
# Run inference directly in the terminal:
./llama-cli -hf ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M

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 ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M

Use Docker

docker model run hf.co/ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M

LM Studio
Jan

vLLM

How to use ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-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": "ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-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/ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M

Ollama
How to use ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF with Ollama:
```
ollama run hf.co/ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M
```
Unsloth Desktop

How to use ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_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": "ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF with Docker Model Runner:
```
docker model run hf.co/ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M
```

Lemonade

How to use ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M

Run and chat with the model

lemonade run user.Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF-IQ1_M

List all available models

lemonade list

Hermes Agent

How to use ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-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 ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_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 ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_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 "ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF:IQ1_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"

Model update: improved reliability, vision and agentic coding

#25

by anm2211 - opened 6 days ago

Discussion

anm2211

IST Austria Distributed Algorithms and Systems Lab org 6 days ago

We’ve updated the model following the community’s feedback. To get the update, you only need to redownload the first GGUF shard. The second shard is unchanged.

We focused on making generation more reliable. Previously, the model could get stuck in repetitive loops or stop before delivering the requested result. In our latest tests, we haven’t encountered either issue. These were limited tests, so we cannot guarantee that either behavior will never occur. Still, the results are encouraging and suggest that both are now less likely. Reports of remaining failure cases would be especially helpful.

We’ve also seen substantial improvements in vision capabilities. Below are two games we built with the updated model to show some of what it can do.

What changed

This update changes which experts are retained. No retained weights were modified relative to the base IQ3_S model. About 25% of the retained expert selection differs from the previous release.

The model remains English-only. With a limited expert budget, we prioritized preserving English capabilities so unfortunately, multilingual capabilities were not preserved.

Benchmark results

SWE-bench Verified accuracy increased from 75.8% to 79.2%, an improvement of 3.4 percentage points in agentic coding compared to 82.8% for the base model. LiveCodeBench v6 accuracy remains essentially unchanged within statistical variation.

Examples built with the updated model

1. A side-scrolling pixel-art cat game

2. A 3D ping-pong game

The ping-pong game supports playing against a bot at different difficulty levels, local split-screen play and online PvP. In the video, I’m controlling both players myself, which explains the rather unusual match!

Give the updated version a try

If you tried the previous release and liked it, we encourage you to download the update. We think you’ll have a smoother, more reliable experience.

If the previous version disappointed you, we’d appreciate you giving it another chance, especially if looping, incomplete responses, or vision performance were the reasons it didn’t work for you.

Thanks to everyone who shared feedback. We’d love to hear how the updated model performs on your tasks, including where it still falls short.

furetosan

6 days ago

WTH is wrong weith you. Now bad model. Beofre the model was good. Why did remove Q4KM fkin

soyalemujica

5 days ago

•

edited 5 days ago

Giving this model a try with a PLE GGUF at FP8, will report results later tomorrow.
Edit: Do not try with a PLE GGUF of FP8, experienced a loop. So now I'm testing with Coder stock GGUF and thus far no loop.

Green-Sky

5 days ago

Good stuff. You might want to update the README.md

leonsarmiento

5 days ago

cool. what inference settings do you recommend?

anm2211

IST Austria Distributed Algorithms and Systems Lab org 5 days ago

cool. what inference settings do you recommend?

Use the default settings recommended by qwen, and we recommend using the model on xhigh reasoning effort.

steamdj

5 days ago

This goes for Qwen3.8-Flash-Next on Strata. Source (https://huggingface.co/unsloth/Qwen3.8-Flash-Next-GGUF):

"sampling": {
"temperature": 1.0,
"top_p": 0.95,
"top_k": 20,
"min_p": 0.0,
"presence_penalty": 0.0,
"repetition_penalty": 1.0
},

soyalemujica

5 days ago

This goes for Qwen3.8-Flash-Next on Strata. Source (https://huggingface.co/unsloth/Qwen3.8-Flash-Next-GGUF):

"sampling": {
"temperature": 1.0,
"top_p": 0.95,
"top_k": 20,
"min_p": 0.0,
"presence_penalty": 0.0,
"repetition_penalty": 1.0
},

Latest strata already passes the default params of the model to the caller, you no longer need to declare these.

Interpause

4 days ago

Giving this model a try with a PLE GGUF at FP8, will report results later tomorrow.
Edit: Do not try with a PLE GGUF of FP8, experienced a loop. So now I'm testing with Coder stock GGUF and thus far no loop.

Interesting, I would like to know why this theoretically happens

xdbrajan3kk

1 day ago

can we get the prompt for;"1. A side-scrolling pixel-art cat game"? it looks great

JayRoss42

about 22 hours ago

yeah this new update broke the model, it doesnt work, please fix this.

anm2211

IST Austria Distributed Algorithms and Systems Lab org about 13 hours ago

@xdbrajan3kk
I actually saw it on X and find it very nice as well. This is the prompt:
"create a simple yet beautiful side scroller where you are a cat, jumping on a city map, collecting fish, html+css+js, pixel art but with details, sound+music, make sure graphics are top notch, fine music, nice gameplay"

@JayRoss42
Could you explain more?

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment