Directory
‹ Back to apps

TokenTally

TokenTally by Harlo is a remote MCP server that counts tokens, truncates text to a budget, chunks text, and estimates model cost. OpenAI tokenizer counts are exact; Claude, Gemini and Llama counts are approximate. Keyless use is about 60 requests/min per IP. Free $0/mo (2,000 calls); Polar Basic $9/mo (50,000) and Pro $15/mo (250,000).

What’s inside

1 MCP server
  • MCP servertokentallyTokenTally by Harlo is a remote MCP server at https://tokentally.mike-tusa.workers.dev/mcp. Tools: count_tokens, truncate_to_budget, chunk_text, estimate_cost. It counts tokens for text in o200k_base and cl100k_base side by side, truncates text to a token budget, chunks text for RAG on paragraph or sentence boundaries with offsets, and estimates cost per model. OpenAI counts (o200k_base, cl100k_base) are exact for those tokenizers. Claude, Gemini and Llama counts are approximate (proxy encoding; real tokenizers differ, often by 10–20%, and newer Claude tokenizers by about 30% or more). Cost estimates use a dated price table (prices_as_of 2026-10-27) and are estimates, not live billing. Stateless: text is not stored or logged; no URL fetching. Max input is 256 KB. A single unbroken run of characters near that cap may hit a CPU timeout (503); normal text at the cap is supported. Pricing: no key, about 60 requests/min per IP. Direct/Polar: Free $0/mo · 2,000 calls/mo; Basic $9/mo · 50,000 calls/mo; Pro $15/mo · 250,000 calls/mo. RapidAPI (public): BASIC $0/mo · 2,000/mo · 10/min; PRO $9/mo · 50,000/mo · 60/min; ULTRA $15/mo · 250,000/mo · 120/min; MEGA $15/mo · 500,000/mo · 120/min. TokenTally is not affiliated with, endorsed by or sponsored by OpenAI, Anthropic, Google or Meta. Model and tokenizer names are used only to describe compatibility and counting methods. Homepage: https://tokentally.mike-tusa.workers.dev Support: digitalpromohub.support@gmail.com