implementation("ai.koog:prompt-tokenizer:1.2.0")
LLM prompt tokenization, model-aware encoders/decoders, token counting, truncation, padding, special-token handling.
