Running LLM Tokenizer Arena & Context Inspector โก Benchmark tokenizers & KV-cache across modern open LLMs