Drop messy retail sales CSVs and get cleaned, profiled, and sanitized data in seconds. Built with FastAPI, Pandas, Neon PostgreSQL, and Next.js—featuring zero-signup Guest Mode (with automatic 30-second database TTL purge), idempotent cloud upserts, and Excel formula-injection protection.
Detect and redact API keys, credentials, and PII locally before sending text to AI. Your text never leaves your browser. The page makes no network requests after it loads.
TokenTab shows what a prompt will cost on Claude, GPT and Gemini before you ship it. Paste a prompt, set reply length and daily requests, and get 10 models ranked by cost per request and per month. It models Batch API discounts and prompt caching, compares two models head to head, and supports USD, PKR and EUR. No signup, no backend: your prompt never leaves your browser. Token counts are labelled exact, close or estimate per model.