Getting Started with NeuroBrowse

Your guide to setting up API keys and understanding the BYOK (Bring Your Own Keys) model

What is BYOK (Bring Your Own Keys)?

BYOK means you use your own API keys from AI providers like OpenAI, Anthropic, Google, and Perplexity. Here's why this matters:

Complete Privacy: Your API keys and summaries stay in your browser. We never see them.
True Cost Transparency: You only pay for what you use, directly to the AI provider.
No Markup: API costs go straight to OpenAI/Anthropic/Google/Perplexity - we don't take a cut.
Your Control: Pick any model, even brand new ones the day they launch.

Privacy First: Your API keys are stored locally in your browser using Chrome's secure storage. NeuroBrowse servers never see your keys, your browsing history, or your summaries. Direct browser-to-AI communication only.

How to Get API Keys

You'll need at least one API key from these providers. We recommend starting with OpenAI or Google (Gemini Flash has a free tier!).

OpenAI (GPT-4o-mini, GPT-4o, GPT-5)

Get your key: platform.openai.com/api-keys

Best for: General summaries, creative writing, code explanation

Cost: ~$0.01 per summary with GPT-4o-mini

Anthropic (Claude Haiku, Claude Sonnet)

Get your key: console.anthropic.com/settings/keys

Best for: Detailed analysis, long documents, technical content

Cost: ~$0.008 per summary with Claude Haiku

Google (Gemini Flash, Gemini Pro)

Get your key: aistudio.google.com/app/apikey

Best for: Fast summaries, free tier available, multi-modal content

Cost: ~$0.005 per summary (or FREE with generous limits!)

Perplexity (Sonar models)

Get your key: perplexity.ai/settings/api

Best for: Real-time web search integration, current events, fact-checking

Cost: ~$0.01 per summary

Understanding Token Costs

AI models charge based on "tokens" - roughly 750 words = 1000 tokens. When you summarize a webpage, you're charged for the input (webpage content) and output (summary).

Typical Summary Costs

Model	Per Summary	100 Summaries/month	500 Summaries/month
Gemini Flash	~$0.005	~$0.50	~$2.50
Claude Haiku	~$0.008	~$0.80	~$4.00
GPT-4o-mini	~$0.01	~$1.00	~$5.00
Perplexity Sonar	~$0.01	~$1.00	~$5.00

Total Monthly Cost: $5 NeuroBrowse subscription + $5-15 API costs = $10-20 total

Compare to ChatGPT Plus: $40/month (and you can't use it on any webpage!)

Which Model Should You Use?

Different models excel at different tasks. Here's our recommendation:

Start with Gemini Flash: Free tier, fast, great for most summaries
Use Claude for: Long academic papers, technical docs, detailed analysis
Use GPT-4o-mini for: Creative content, code, conversational summaries
Use Perplexity for: News articles, current events, when you need citations

The beauty of NeuroBrowse? Switch between them with one click. Try different models on the same page and compare!

Setup Wizard

When you first install NeuroBrowse, our setup wizard will:

Walk you through subscribing ($5/month)
Guide you to get API keys (we link directly to each provider)
Securely store your keys in Chrome's local storage
Test each key to make sure it works
Set your preferred default model

Total setup time: 5-10 minutes. Then you're ready to start summarizing!

Managing Your Costs

Each AI provider has a dashboard where you can:

Set monthly spending limits
View your usage in real-time
Get billing alerts
See cost breakdowns by model

We recommend setting a $20/month limit when you first start. You'll likely use much less, but it gives you peace of mind.

Back to Home Subscribe - $5/month