About PromptUtils
What is PromptUtils?
PromptUtils is a collection of free, browser-based tools built by a software engineer working in machine learning and NLP. These tools solve real problems I encounter in my own research and development work, and I'm sharing them because they might help you too.
Every tool runs entirely in your browser, which means:
- ✓ Your data never leaves your machine
- ✓ No signup or authentication required
- ✓ Lightning-fast performance (no network latency)
- ✓ Full privacy by design
- ✓ Works offline after initial load
Why I Built PromptUtils
During my work in ML and NLP, I found myself constantly needing quick utilities to solve repetitive problems. Rather than switching between different tools and websites, I built PromptUtils as a single place to have everything I need:
- Quickly count tokens in text during prompt development
- Calculate API costs to estimate research budgets
- Generate structured prompts for consistency
- Test and debug regex patterns
- Format and parse data for model input
- Analyze and augment training data
- Understand model parameters (temperature, top-k, etc.)
- Estimate computational requirements for experiments
These tools have made my work more productive, and I'm sharing them because they might help you too. No signup. No fees. No data collection. Just practical tools that work.
My Philosophy
Privacy First
I never store your data or results. Everything processes in your browser—you maintain complete control.
Always Free
No premium tiers, no paywalls, no feature gating. Built for the community, not for profit.
Practical Focus
Every tool solves a real problem I encounter in ML/NLP work. Fast, lightweight, no bloat.
Open Source
Code is transparent, auditable, and available for modification and contribution.
Tool Categories
💰 Cost & Tokenization
Calculate LLM API costs, count tokens, manage context windows, and estimate fine-tuning expenses across all major providers.
📊 Data & Training
Generate augmented training data, preprocess text, analyze class balance, and plan transformer fine-tuning campaigns.
⏰ Time & Utilities
Work with Unix timestamps, cron expressions, JSON validation, regex testing, and Base64 encoding.
🧠 Advanced
Specialized tools for prompt template generation, text preprocessing, and model behavior analysis.
Common Use Cases
For ML Engineers
- Estimate BERT fine-tuning costs and GPU hours
- Generate augmented training data with EDA techniques
- Analyze class balance and calculate sample weights
- Preprocess text datasets at scale
For LLM/Prompt Engineers
- Count tokens and manage context windows
- Estimate API costs before running expensive queries
- Compare prompt variations and their token impact
- Generate structured prompt templates
- Understand sampling parameters (temperature, top-k)
For Researchers
- Benchmark model tokenization across providers
- Analyze text preprocessing pipelines
- Generate diverse training samples from limited data
- Estimate computational requirements for experiments
For Developers
- Format and validate JSON data
- Test and debug regular expressions
- Convert and work with timestamps
- Generate and manage cron schedules
- Encode and decode Base64 data
Technology Stack
Frontend: Vanilla JavaScript, HTML5, CSS3 (no heavy frameworks—just what you need)
Processing: Client-side only (runs in your browser)
Data Storage: None (no servers, no databases)
Analytics: Google Analytics (optional, you can opt out)
Hosting: Static files (CDN-friendly, lightning fast)
Open Source & Community
PromptUtils is fully open source. Once the repository is available, you can:
- View and contribute to the source code
- Report bugs and suggest features
- Contribute code improvements
- Fork the project for your own use
- Deploy it on your own servers
Privacy & Data Security
Your privacy is fundamental to how we operate:
- No User Accounts: We don't collect or store user information
- No Data Transmission: All processing happens in your browser—your data never leaves your machine
- No Cookies (except ads): We use Google Analytics and AdSense cookies only (you can opt out)
- HTTPS Only: All communication is encrypted
- Open Code: You can audit the source to verify our claims
For details, see our Privacy Policy.
Roadmap & Future
We're constantly adding new tools and improving existing ones based on community feedback. Areas we're exploring:
- More LLM provider support (Claude, Gemini, custom models)
- Advanced prompt optimization and comparison tools
- Integration with popular IDEs and editors
- Desktop applications for offline use
- Collaboration features for teams
- Export and visualization tools
About Me
I'm a software engineer specializing in machine learning and NLP. I built PromptUtils to solve problems I encounter in my own research and development work. If you find these tools useful or have ideas for improvements, I'd love to hear from you.
Get in touch:
- Email: jcbolo72012@gmail.com
- Portfolio: View my projects and research
- Contact: Use the contact form for feedback and suggestions