Resources // TERSL

Compare your Code

Paste your JSON or YAML and see what it really costs. Your data converts to TERSL side by side — same information, live token counts, and the exact savings you'd bank on every model call. Runs entirely in your browser; nothing is uploaded. This is an illustrative encoder — for the normative mapping, read the spec. Want to go deeper than counting? Explore TERSL live.

Examples //

0 tok

TERSL · output

0 tok

—

fewer tokens than JSON

JSON

— tok

YAML

— tok

TERSL

— tok

Token counts are estimates (≈ 1 token per 3.8 characters, GPT-class tokenizers). JSON measured pretty-printed; YAML via canonical dump. Your exact savings depend on the model's tokenizer and your data's shape.