About Evvl
A privacy-first tool for comparing AI model outputs
What is Evvl?
Evvl is a simple, privacy-focused tool that lets you compare outputs from different AI models side by side. Enter a single prompt, select which AI providers you want to use, and instantly see how different models respond to the same input.
Whether you're evaluating which AI model works best for your use case, testing prompt variations, or just curious about how different models approach the same question, Evvl makes it easy to compare and contrast their outputs.
How It Works
- Add your API keys: Visit the Settings page and add API keys from OpenAI, Anthropic, and/or OpenRouter. These are stored locally in your browser.
- Enter a prompt: Type the prompt you want to test on the main Eval page.
- Generate outputs: Click “Generate Outputs” to send your prompt to all configured AI providers simultaneously.
- Compare results: View the responses side by side, along with token counts and latency metrics for each model.
Privacy First
Evvl takes your privacy seriously. Here's what makes it different:
- API keys protected: Your API keys are automatically redacted from all server logs. They are never stored in any database. Prompts and AI responses are never logged at all.
- Transparent proxy: Your requests are routed through our server (required for OpenAI compatibility). We maintain basic operational logs (provider, model, error types) for debugging, but no personal or sensitive data.
- No tracking: We don't track your usage, collect analytics, or store any information about your prompts or results.
- Open and transparent: The tool is straightforward and honest about how it works—no hidden features or data collection.
Free to Use
Evvl itself is completely free to use. You only pay for the API calls made to the AI providers (OpenAI, Anthropic, OpenRouter) based on your usage with them. There are no subscription fees, no hidden costs, and no premium tiers.