REAL AI USAGE RANKING
WHO REALLY USES AI?
Measure real AI usage.
See who uses the most.
See which LLMs people actually use.
No claims · No self-reporting · Just usage · Updated from completed UTC days
REAL TOKEN USAGE RANKING
Who is using AI the most?
Ranked by verified aggregate Token usage.
LLM RANK
WHICH LLMS ARE PEOPLE ACTUALLY USING?
MADRank aggregates real user Token usage by model.
LLM Rank is derived from aggregated real user usage — it measures volume, not capability.
HOW MADRANK WORKS
Install MADRank into DeepSeek Harness (DSH).
Use AI normally. MADRank measures supported Token usage automatically.
Completed aggregate usage data is synced to MADRank.
Your usage contributes to your public rank. Your model usage contributes to the LLM Rank.
THE NEUTRAL USAGE LAYER FOR AI
DSH FIRST. EVERY AI CLIENT NEXT.
The DSH plugin is the official client — zero-config, always on. The same node also reads every other AI client on your machine: Run madrank-node doctor anytime to see per-client health.
WE RANK USAGE, NOT CONTENT.
MADRank only needs aggregate usage signals required for ranking.
No prompt content · No response content · No files · No API keys · No self-reported Token counts
Real usage · Minimum data · Public rank
READ PRIVACY →GET YOUR RANK.
Install MADRank.
Use AI normally.
See where you stand.