Instructions to use Valnivo-labs/valnivo-copilot-1-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 Valnivo-labs/valnivo-copilot-1-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 Valnivo-labs/valnivo-copilot-1-gguf:Q8_0
# Run inference directly in the terminal:
llama cli -hf Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Valnivo-labs/valnivo-copilot-1-gguf:Q8_0
# Run inference directly in the terminal:
llama cli -hf Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

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 Valnivo-labs/valnivo-copilot-1-gguf:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

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 Valnivo-labs/valnivo-copilot-1-gguf:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

Use Docker

docker model run hf.co/Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

LM Studio
Jan

vLLM

How to use Valnivo-labs/valnivo-copilot-1-gguf with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Valnivo-labs/valnivo-copilot-1-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": "Valnivo-labs/valnivo-copilot-1-gguf",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

Ollama
How to use Valnivo-labs/valnivo-copilot-1-gguf with Ollama:
```
ollama run hf.co/Valnivo-labs/valnivo-copilot-1-gguf:Q8_0
```
Unsloth Desktop

How to use Valnivo-labs/valnivo-copilot-1-gguf with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

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": "Valnivo-labs/valnivo-copilot-1-gguf:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Valnivo-labs/valnivo-copilot-1-gguf with Docker Model Runner:
```
docker model run hf.co/Valnivo-labs/valnivo-copilot-1-gguf:Q8_0
```

Lemonade

How to use Valnivo-labs/valnivo-copilot-1-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

Run and chat with the model

lemonade run user.valnivo-copilot-1-gguf-Q8_0

List all available models

lemonade list

Hermes Agent

How to use Valnivo-labs/valnivo-copilot-1-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 Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

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 Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Valnivo-labs/valnivo-copilot-1-gguf with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

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 "Valnivo-labs/valnivo-copilot-1-gguf:Q8_0" \
  --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"

Valnivo Copilot 1.6 for Ollama

Valnivo Copilot, the optional model of Valnivo's question box "Ask Valnivo", by Valnivo Labs, as a GGUF file for Ollama — the free helper: it runs on the person's own computer, outside the browser, and is faster there than the browser build on WebGPU. Valnivo's page talks to it at 127.0.0.1:11434, so what the person asks stays on their computer.

The same model as Valnivo-labs/valnivo-copilot-1 at tag v1.6 — Qwen2.5-1.5B-Instruct (© Alibaba Cloud, Apache-2.0) fine-tuned with LoRA to route a question to an entry of Valnivo's own library and to word a reply from figures the app has already computed. That card says how it was trained (Valnivo's own library and reply templates only; no person's data) and what it may and may not do. It gives no advice: Valnivo's code refuses a request for a verdict before any model runs, and shows only library ids and replies that pass its checks.

Use

ollama pull hf.co/Valnivo-labs/valnivo-copilot-1-gguf:Q8_0

Then let Valnivo's site reach it — on a Mac launchctl setenv OLLAMA_ORIGINS "<the site's address>", elsewhere the OLLAMA_ORIGINS variable — restart Ollama, and choose In the free helper in Ask Valnivo. The panel shows both commands with a copy button.

Files

File	Size
`valnivo-copilot-1-Q8_0.gguf`	1.65 GB	8-bit, the one Valnivo asks for

Q8_0 is the one Valnivo asks for: 8-bit, practically lossless, where the browser build is held to 4-bit by WebGPU. Converted from the fused model with llama.cpp's convert_hf_to_gguf.py, quantised with llama-quantize.

Measured through Ollama (held out, never trained on)

Set	Result
Writing, 128 questions, 8 languages	128/128 pass the app's checks
eval4, 200 questions written blind	94% right, 0 of 24 advice questions leaked
Casual questions, 80, written blind	93% right, 0 of 8 advice questions leaked
History and money lent between two users, 72	99% right, 0 of 8 advice questions leaked
Round 6 (debts, loans, salary over time, the situation, discrepancies), 240	96% right, 0 of 8 advice questions leaked
Privacy notice and terms, 48 questions	79% right
eval3, 39 questions	38/39 right, 0 of 9 advice questions leaked, 9/9 off-topic → none
Actions, 41 instructions	39/41 right, no wrong action offered

About 0.35 s to route a question on an Apple M3. Q5_K_M and Q4_K_M were built from the same model and measured too; neither did better, so only Q8_0 is published.

These are the same questions and facts the browser build is measured on, sent the way the app sends them: greedy, with no structured output — the app keeps only a reply that is one of its library's ids. (Ollama's structured output was measured too, and cost 5 to 11 points: the model was made to open with a quote mark it never saw in training.)

Versions

This repository carries the current version only — Ollama pulls by quantisation, not by tag. Earlier versions are tags of Valnivo-labs/valnivo-copilot-1.