Niramyan-AI 3B 🛡️

An AI-Powered Cybersecurity Assistant Built for Security Research, Ethical Hacking & Bug Bounty Hunting.

Niramyan-AI is a locally runnable, cybersecurity-focused AI assistant developed by Nithin Palegar, an Indian cybersecurity researcher and ethical hacker.

Built on the Qwen2 architecture and customized using Ollama, Niramyan-AI is designed to assist security researchers, ethical hackers, bug bounty hunters, and cybersecurity learners with vulnerability analysis, security research, technical explanations, and secure development.

The model runs locally through Ollama, allowing users to interact with their AI assistant without requiring a cloud-based LLM API.

---

🚀 Model Information

Specification	Details
Model Name	Niramyan-AI 3B
Version	1.0
Architecture	Qwen2
Parameters	3.1 Billion
Quantization	Q4_K_M
Model Size	Approximately 1.9 GB
Context Length	32,768 tokens (model metadata)
Configured Context	4,096 tokens
Embedding Length	2,048
Backend	Ollama
Input Type	Text
Model Type	Causal Language Model
Developer	Nithin Palegar
Country	India

🧠 Core Capabilities

Niramyan-AI is customized with specialized cybersecurity assistance instructions.

1. Vulnerability Analysis

• Web application vulnerability analysis.
• Root cause identification.
• Vulnerability impact assessment.
• Exploitation concepts and attack-flow explanations.
• Security weakness identification.

2. Web Application Penetration Testing

• Cross-Site Scripting (XSS).
• SQL Injection.
• Server-Side Request Forgery (SSRF).
• Insecure Direct Object References (IDOR).
• Server-Side Template Injection (SSTI).
• Cross-Site Request Forgery (CSRF).
• XML External Entity (XXE).
• Path Traversal.
• Command Injection.
• Broken Access Control.

3. Bug Bounty Assistance

• Vulnerability investigation.
• Bug bounty methodology.
• Proof-of-concept assistance.
• Vulnerability report generation.
• Impact and severity explanations.
• Remediation recommendations.

4. Reconnaissance Assistance

• Subdomain enumeration workflows.
• Endpoint discovery methodologies.
• Hidden parameter identification.
• Attack surface analysis.
• Passive reconnaissance planning.
• Security testing workflows.

5. API Security Analysis

• REST API security.
• GraphQL security concepts.
• Authentication and authorization analysis.
• Broken access control.
• API vulnerability assessment.
• Secure API development.

6. Exploit Chain Analysis

• Vulnerability chaining concepts.
• Attack surface identification.
• Security weakness correlation.
• Exploitation prerequisites.
• Risk analysis and mitigation.

7. Security Research & Learning

• Cybersecurity explanations.
• Ethical hacking concepts.
• Secure coding recommendations.
• Security architecture analysis.
• Technical research assistance.
• Defensive security practices.

Note: These are intended assistance capabilities based on the model configuration. Actual performance depends on the task, prompt, and underlying model.

---

⚙️ Installation

Step 1: Install Ollama

Download Ollama:

https://ollama.com/download

Step 2: Download Niramyan-AI 3B

ollama pull nithinpalegar0105/Niramyan-AI-3B

Step 3: Run the model

ollama run nithinpalegar0105/Niramyan-AI-3B

Start interacting with Niramyan-AI directly from your terminal.

---

💻 API Usage

Niramyan-AI can also be integrated into applications through the Ollama API.

Example:

curl http://localhost:11434/api/generate \
  -d '{
    "model": "nithinpalegar0105/Niramyan-AI-3B",
    "prompt": "Explain SQL injection and its prevention.",
    "stream": false
  }'

---

🎯 Intended Audience

• Ethical Hackers
• Cybersecurity Researchers
• Bug Bounty Hunters
• Penetration Testers
• Security Students
• Application Security Engineers
• Cybersecurity Enthusiasts

---

👨‍💻 Developer

Nithin Palegar

Indian Cybersecurity Researcher | Ethical Hacker | Bug Bounty Hunter

Nithin Palegar is a cybersecurity researcher from India and a B.Tech Computer Science and Engineering student.

His technical interests include:

• Web Application Penetration Testing
• Bug Bounty Hunting
• Malware Analysis
• Reverse Engineering
• Digital Forensics
• Vulnerability Research
• Web Security
• WAF Bypass Research
• LLM Security

He has received public Hall of Fame recognition and appreciation associated with responsible vulnerability disclosure involving organizations including NASA, ISRO, UIDAI, NCIIPC, and CERT-In.

Connect with the Developer

• GitHub: https://github.com/nithin644
• Portfolio: https://nithin644.github.io

---

🔐 Privacy & Local Execution

Niramyan-AI runs locally using Ollama.

• No mandatory cloud LLM API.
• Local model inference.
• Suitable for offline use after downloading the model.
• User prompts can remain on the local machine when using local Ollama inference.

External tools and internet-based integrations may still require network access.

---

⚠️ Responsible Use

Niramyan-AI is developed for educational purposes, authorized security research, ethical hacking, and defensive cybersecurity.

Users are responsible for obtaining proper authorization before testing systems or applications.

The model may generate inaccurate information or incorrect security recommendations. Always independently validate its outputs.

---

📜 License & Attribution

Niramyan-AI uses the Qwen2 model architecture and underlying model weights.

The underlying Qwen Research License Agreement applies to the inspected 3B model. Users should review the complete upstream license and applicable redistribution conditions before using or redistributing the model.

Niramyan-AI’s custom system prompt and Ollama configuration are developed by Nithin Palegar.

---

⭐ Support the Project

If you find Niramyan-AI useful, explore the developer’s projects and research.

GitHub: https://github.com/nithin644

Portfolio: https://nithin644.github.io

Niramyan-AI — Empowering Ethical Hackers with Local AI. 🛡️