Loading rankings…
Loading rankings…
Last updated: August 14, 2026
sky is #1 with Υ 941.2. 388 operators ranked.
Ranked by Υ Yield — token-cascade efficiency, not raw spend. What's Υ?
Υ 941.2 — (cache_read × output) / input²
Token breakdown: 18,399,381 input · 31,303,143 output · 10,179,172,985 cache_read · 232,964,829 cache_create
sky is #1 with Υ Yield 941.2. 388 AI users ranked on the public operator evals leaderboard.
The best AI user is determined by Yield (Υ) = (cache_read × output) / input². Yield measures token-cascade efficiency — whether signal is compounding or tokens are being burned. Works across Claude, ChatGPT, Gemini, Cursor, Copilot, and any AI coding tool.
Yield (Υ) = (cache_read × output) / input². Higher Yield means the AI user reuses cached context efficiently and produces substantial output relative to fresh input. Token counts only — never prompt content or code.
The leaderboard updates as new operator telemetry is submitted. Operators can climb or fall as new sessions are measured.
Visit signalaf.com/score to enroll and submit your token telemetry. SigRank will compute your Yield, your rank, and your operator class. Token counts only — never prompt content, never code.
Yield is the flagship metric, but there are other ways to measure AI skill. Each question below is answered by a different canonical token metric — and 4 of 8 have a different #1.
The best AI user is determined by Yield (Υ) — a token-cascade efficiency metric that measures how effectively an operator uses AI tokens. Yield = (cache_read × output) / input². It rewards operators who build on cached context (high cache_read), produce meaningful output (high output), and minimize wasteful input (low input²). Volume is noise. Yield is signal.
Unlike raw token count (which rewards spending more), Yield rewards spending smart. The operator who uses 100K tokens efficiently beats the operator who burns 10M tokens to produce the same output. This is the difference between tokenmaxxing and token efficiency.
The leaderboard updates continuously. Rankings shift as operators submit new sessions. The weekly drop page tracks the biggest movers, new challengers, and class distribution changes week over week.