Instructions to use gorse-io/Qwen3-Reranker-8B-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 gorse-io/Qwen3-Reranker-8B-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 gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf gorse-io/Qwen3-Reranker-8B-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 gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_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 gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_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 gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use gorse-io/Qwen3-Reranker-8B-GGUF with Ollama:
```
ollama run hf.co/gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M
```
Unsloth Desktop

How to use gorse-io/Qwen3-Reranker-8B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf gorse-io/Qwen3-Reranker-8B-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": "gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use gorse-io/Qwen3-Reranker-8B-GGUF with Docker Model Runner:
```
docker model run hf.co/gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M
```

Lemonade

How to use gorse-io/Qwen3-Reranker-8B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen3-Reranker-8B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use gorse-io/Qwen3-Reranker-8B-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 gorse-io/Qwen3-Reranker-8B-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 gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use gorse-io/Qwen3-Reranker-8B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf gorse-io/Qwen3-Reranker-8B-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 "gorse-io/Qwen3-Reranker-8B-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"

Qwen3-Reranker-8B

GGUF conversion of Qwen/Qwen3-Reranker-8B for dedicated query/document reranking, not text generation or general-purpose embedding vectors. Available files:

Qwen3-Reranker-8B-BF16.gguf
Qwen3-Reranker-8B-Q8_0.gguf
Qwen3-Reranker-8B-Q4_K_M.gguf

Run with llama.cpp

Use a recent llama.cpp build with Qwen3 reranker conversion and server support:

llama-server -hf gorse-io/Qwen3-Reranker-8B-GGUF:Q8_0 \
    --embedding --reranking --pooling rank --host 127.0.0.1 --port 8080

For the 4-bit variant, use -hf gorse-io/Qwen3-Reranker-8B-GGUF:Q4_K_M in the same command. For a local file, replace -hf ... with -m Qwen3-Reranker-8B-Q8_0.gguf or -m Qwen3-Reranker-8B-Q4_K_M.gguf. The conversion pipeline builds llama-quantize, not llama-server; build or install the server separately.

curl http://127.0.0.1:8080/v1/rerank \
    -H 'Content-Type: application/json' \
    -d '{"query":"What is the capital of China?","documents":["Beijing is the capital of China.","Paris is the capital of France."],"top_n":2}'

/reranking, /rerank, and /v1/reranking are aliases. The endpoint scores query/document pairs; it is not a chat-completions endpoint. See the llama.cpp server documentation.

Conversion notes

Keep the original downloaded source README.md alongside the weights and configuration. llama.cpp detects Qwen3 rerankers using the source directory name or original README, emits RANK pooling, and extracts the yes/no rows into cls.output.weight. The pipeline's temporary directory names use underscores, so preserving the upstream README matters. Do not replace it with this output model card before conversion, especially when using a local source directory.

CPU conversion is supported, but budget RAM for conversion/quantization and disk for the source weights, Hugging Face cache, and BF16, Q8_0, and Q4_K_M outputs. The larger models need substantially more memory and storage; --keep retains the downloads and all three output files.