Analyze your prompt. See what it costs. Make it cheaper.

Paste any prompt to get token counts, per-model costs, and compressed versions at all 3 levels — then share it with a permanent URL.