This video demonstrates how to set up NanoBrowser AI and connect it to multiple LLMs in a single workflow:
1. Direct OpenAI integration with your API key
2. OpenRouter to access DeepSeek, Gemini, and Anthropic Claude using one API
3. A local Dolphin-Mistral model exposed on an OpenAI-compatible endpoint

The goal is a reliable, privacy-aware multi-LLM setup you can switch between for quality, latency, and cost—without changing your browser.

🖥️ What you will learn
1. Install and configure NanoBrowser AI
2. Configure providers: OpenAI, OpenRouter, and a local OpenAI-compatible server
3. Connect multiple LLMs through OpenRouter (DeepSeek, Gemini, Claude)
4. Add a local Dolphin-Mistral model (e.g., via LM Studio or Ollama)
5. Validate each connection with request examples

🔗 Links
1. NanoBrowser: https://nanobrowser.ai/
2. GitHub (NanoBrowser): https://github.com/nanobrowser/nanobrowser
3. OpenAI: https://platform.openai.com/
4. OpenRouter: https://openrouter.ai/

✅ Prerequisites
1. API keys: OPENAI_API_KEY, OPENROUTER_API_KEY
2. A local OpenAI-compatible server running Dolphin-Mistral (e.g., LM Studio or Ollama)
3. Base URLs you will configure in NanoBrowser:
o OpenAI: https://api.openai.com/v1
o OpenRouter: https://openrouter.ai/api/v1
o Local: http://localhost:11434/v1 (adjust for your runtime)

************Never expose API keys on screen. Use environment variables or a secrets manager and revoke/regenerate keys if they’re ever leaked.*******

#NanoBrowser #LLM #OpenAI #OpenRouter #DeepSeek #Gemini #Anthropic #Mistral #AIAgents #WebAutomation #Privacy #DFIR #KisoSecure

If you want, I can tailor this with your brand voice (Kiso Secure), add a pinned comment template, or swap in exact local runner instructions you used.