Instructions to use ModernoTech/Nova-Spark with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use ModernoTech/Nova-Spark 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 ModernoTech/Nova-Spark:Q4_0
# Run inference directly in the terminal:
llama cli -hf ModernoTech/Nova-Spark:Q4_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ModernoTech/Nova-Spark:Q4_0
# Run inference directly in the terminal:
llama cli -hf ModernoTech/Nova-Spark:Q4_0

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 ModernoTech/Nova-Spark:Q4_0
# Run inference directly in the terminal:
./build/bin/llama cli -hf ModernoTech/Nova-Spark:Q4_0

Use Docker

docker model run hf.co/ModernoTech/Nova-Spark:Q4_0

Llama new
LM Studio
Jan

vLLM

How to use ModernoTech/Nova-Spark with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ModernoTech/Nova-Spark"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ModernoTech/Nova-Spark",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ModernoTech/Nova-Spark:Q4_0

Ollama
How to use ModernoTech/Nova-Spark with Ollama:
```
ollama run hf.co/ModernoTech/Nova-Spark:Q4_0
```
Unsloth Desktop

How to use ModernoTech/Nova-Spark with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ModernoTech/Nova-Spark:Q4_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": "ModernoTech/Nova-Spark:Q4_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use ModernoTech/Nova-Spark with Docker Model Runner:
```
docker model run hf.co/ModernoTech/Nova-Spark:Q4_0
```

Lemonade

How to use ModernoTech/Nova-Spark with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ModernoTech/Nova-Spark:Q4_0

Run and chat with the model

lemonade run user.Nova-Spark-Q4_0

List all available models

lemonade list

Hermes Agent

How to use ModernoTech/Nova-Spark with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ModernoTech/Nova-Spark:Q4_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 ModernoTech/Nova-Spark:Q4_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use ModernoTech/Nova-Spark with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ModernoTech/Nova-Spark:Q4_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 "ModernoTech/Nova-Spark:Q4_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"

Nova Spark 1.0

Nova Edge · Version 1.0

Nova Spark 1.0 — Moderno Tech Responsive conversational AI for local edge devices.

Responsive conversational model for edge and browser runtimes, with compact deployment and low-latency local inference.

Repository · Download · Moderno Tech

Nova Spark 1.0

Nova Spark 1.0 is a Moderno Tech product in the Nova Edge family.

Moderno Tech / Nova Identity

Nova Spark 1.0 is presented as an independent product of the Moderno Tech Nova platform. Its commercial identity, naming, versioning, documentation and repository belong to Moderno Tech.

Moderno Tech does not claim that these weights were trained from scratch. NATIVE, FINE-TUNED and DISTILLED variants will be published in future versions with their own training documentation.

What Nova brings

Modelo conversacional compacto para inferencia local en navegadores y dispositivos edge. Está orientado a respuestas ágiles, asistencia práctica y ejecución eficiente, incluso en entornos con recursos limitados.

Core specifications

Field	Value
Product	Nova Spark 1.0
Organization	Moderno Tech
Family	Nova Edge
Status	Published and verified distribution
Base parameters	0.8B (800M)
Commercial version	`1.0`
Architecture identity	Nova-branded product, technical base documented below
Format	GGUF
Quantization	Q4_0
File	`Nova-Spark-1.0-Q4_0.gguf`
Expected bytes	563036064
SHA-256	`57d1997790d1744fba5b40a7317df71ea5e2acee28c47e78f0cce39c0703f8cf`

Key strengths

Conversación conversacional local
Respuestas ágiles
Asistencia práctica
Inferencia compacta y eficiente

Downloads

Variant	File	Format	Size	Status
Stable	Nova-Spark-1.0-Q4_0.gguf	GGUF Q4_0	563036064 bytes	Published and verified

Each download points to the exact published artifact in this repository.

Installation and usage

Use a GGUF-compatible runtime such as llama.cpp, Ollama, or an equivalent local inference engine, and keep the original artifact integrity.

Example installation reference:

llama-cli -hf ModernoTech/Nova-Spark --conversation

This command must be tested with the exact runtime and model format before claiming full compatibility. Follow the original author's recommended parameters whenever available.

Technical documentation

Format: GGUF
Quantization: Q4_0
File: Nova-Spark-1.0-Q4_0.gguf
Expected bytes: 563036064
SHA-256: 57d1997790d1744fba5b40a7317df71ea5e2acee28c47e78f0cce39c0703f8cf
Runtime family: llama.cpp

Capabilities and use cases

Conversación conversacional local
Respuestas ágiles
Asistencia práctica
Inferencia compacta y eficiente

Limitations

This repository is a branded distribution of a verified original artifact.
No weights, architecture, tokenizer, prompts or inference parameters were altered.
Supported runtime and application scenarios must be validated with the exact hardware and browser/runtime in use.

Technical provenance and licensing

Field	Value
Product identity	Moderno Tech Nova
Published artifact	`Nova-Spark-1.0-Q4_0.gguf`
Base model	`Qwen/Qwen3.5-0.8B`
Original repository	`ggml-org/Qwen3.5-0.8B-GGUF`
Original artifact filename	`Qwen3.5-0.8B-Q4_0.gguf`
Quantization	`Q4_0`
Expected bytes	`563036064`
SHA-256	`57d1997790d1744fba5b40a7317df71ea5e2acee28c47e78f0cce39c0703f8cf`
Original license	`Apache License 2.0`

The original model and authors retain their corresponding authorship and licensing rights. Moderno Tech preserves attribution and legal notices as required by the original license.

Evolution model

Nova versions will evolve without changing product identity or repository:

1.0: initial branded distribution based on verified original weights.
1.1: incremental fine-tuning and documented behavior improvements.
1.5: significant adaptation and training changes.
2.0: new engine generation with Nova-native weights.
3.0: mature next-generation Nova.

A version containing fine-tuning must have new hashes, training documentation, provenance and model format details. A version with unchanged weights must never be presented as native training.

Version policy

Nova Spark 1.0 is the first published Nova version.
Future improvements must create a new version number, even if the base architecture remains related.
Quantizations are distribution variants of the same commercial version.

Institutional footer

MODERNO TECH

Nova AI

Intelligence. Engineering. Innovation.

Moderno Tech: https://www.moderno.com.ar
Hugging Face organization: https://huggingface.co/ModernoTech
Repository: https://huggingface.co/ModernoTech/Nova-Spark

Training and Distillation

This is the initial Moderno Tech Nova publication.

Training status: foundation
Training performed by Moderno Tech for this publication: not yet documented
Fine-tuning performed: not yet documented
Distillation performed: not yet documented

This repository may be updated in the future with the corresponding training checkpoint, LoRA, dataset, configuration and evaluation records.

Teacher model

Not publicly documented for this version.

Student model

The published artifact is verified and kept intact. Its weight origin is documented in the Technical provenance and licensing section.

Weight provenance

The published GGUF artifact was redistributed and rebranded as a Moderno Tech Nova product. No weight tensor was modified.

Metric, training configuration and teacher data are intentionally not declared because they are not available for this version.

Future evolution

Future versions may include:

1.1: incremental fine-tuning
1.5: training-based adaptation
2.0: distillation or weights developed by Moderno Tech
3.0: future Nova-native generation

Downloads last month: -

GGUF

Model size

0.8B params

Architecture

qwen35

Hardware compatibility

4-bit