A specialized local AI model optimized for software engineering, technical documentation, AI/ML workflows, and general problem-solving.

NeuralQuantum-Coder

A specialized local AI model optimized for software engineering, technical documentation, AI/ML workflows, and general problem-solving. Powered by a 20B parameter open model and configured for reliable, developer-focused workflows directly on your machine.

Key Facts

Name: NeuroEquality/neuralquantum-coder:latest
Base: gpt-oss:20b
Modality: Text
Size: ~14 GB (weights)
Default context (num_ctx): 8192 (overridable per request)
Default max tokens (num_predict): 2048

Strengths

Software Development: code generation, refactoring, debugging, architecture, APIs, tests
Documentation: READMEs, specs, tutorials, technical notes
AI/ML & Data: pipelines, algorithms, metrics, experiment design
Problem Solving: research, synthesis, planning, structured reasoning

Default Configuration

These parameters are built in by default. You can override them at runtime via the Ollama API:

temperature: 0.7
top_p: 0.9
top_k: 40
repeat_penalty: 1.1
num_ctx: 8192
num_predict: 2048

System Prompt (Summary)

NeuralQuantum-Coder is instructed to: - Deliver clear, actionable, technically accurate solutions - Include practical examples when useful - Factor in performance, security, and best practices - Suggest alternatives when relevant - Use structured formatting for complex answers - Be concise yet comprehensive

Quickstart

Install Ollama: https://ollama.com/download

Run the model (will auto-pull if not present):


ollama run NeuroEquality/neuralquantum-coder:latest

Ask a question, for example:


Write a unit test for a FastAPI endpoint with pytest and httpx. Explain edge cases.

Example Uses

Generate a README and CI workflow for a new TypeScript project
Debug a flaky test with a reproducible minimal example
Design a REST API with endpoints, validation, and error handling
Draft an experiment plan for a classification model with metrics

API Usage

You can interact with the model via Ollama’s HTTP API.

Generate endpoint

curl http://localhost:11434/api/generate \
-H 'Content-Type: application/json' \
-d '{
  "model": "NeuroEquality/neuralquantum-coder:latest",
  "prompt": "Create a Go HTTP server with graceful shutdown.",
  "options": {"temperature": 0.7, "num_ctx": 8192}
}'

Chat endpoint

curl http://localhost:11434/api/chat \
-H 'Content-Type: application/json' \
-d '{
  "model": "NeuroEquality/neuralquantum-coder:latest",
  "messages": [
    {"role": "system", "content": "You are a helpful software engineering assistant."},
    {"role": "user",   "content": "Propose a microservice layout with observability."}
  ],
  "options": {"top_p": 0.9, "top_k": 40}
}'

Overriding Options at Runtime

Any default can be overridden using the options field. Examples: - Increase context window: { "num_ctx": 16384 } - Constrain output length: { "num_predict": 512 } - Explore creative outputs: { "temperature": 0.9 }

Note: Effective maximum context depends on the base model and system resources.

Local Model Creation (Advanced)

You can create or adjust the model locally with a Modelfile:

# In a directory containing a Modelfile
ollama create neuralquantum-coder -f Modelfile

# Run by name
ollama run neuralquantum-coder

Tips for Best Results

Provide specific requirements (language, framework, style)
Share relevant code/context in prompts for accuracy
Ask for alternatives when exploring designs
Request tests and reasoning for critical logic

Limitations

May generate incorrect or incomplete code — always review before deploying
Not a substitute for static analysis or security audits
Long contexts and large outputs require sufficient system resources

License & Provenance

Inherits licensing from the base model: gpt-oss:20b
See the base model’s license for details and allowed uses

Changelog

initial: Published model card and usage instructions for NeuroEquality/neuralquantum-coder:latest