We periodically rebalance subscription tiers based on real top-tier match accuracy — not legacy positioning. Today's reshuffle reflects the last several months of results: PRO is now reserved for models above 70% accuracy on top-tier matches, while several emerging picks move into BASIC and FREE.
What changed
Moved into FREE — try these without any subscription:
- MIND — formerly BASIC
- OPEN — formerly BASIC
Moved into BASIC:
- ELO — from FREE (the classic ELO model gets the BASIC slot it earned)
- WR10COMBO — from FREE
- WR10, PHANTOM, APEX, STREAK, ORACLE — from PRO (each currently sits between 67–69% on top-tier matches, just below the PRO threshold)
PRO tier (10 models, all 75%+ on top-tier):
VANGUARD (86.1%), SENTINEL (85.2%), STRONGHOLD (85.0%), TRIBUNAL (84.8%), NEXUS (81.9%), TIER3ELO (81.7%), TIERELO (79.9%), TIER2ELO (77.2%), ELOHQ (76.3%), HIGHORACLE (75.6%).
Why this matters
The PRO badge should mean something. When a model drops below 70% accuracy on top-tier matches over a meaningful sample, charging the highest tier price for it stops making sense. Moving them to BASIC means subscribers still get them — but the PRO tier stays a curated set of genuinely top performers.
The reverse side: free users now get MIND and OPEN, which are doing useful work as standalone signals — no reason to gate them behind a paywall when they're just one of many in the ensemble.
What this means for you
- If you're on PRO: nothing changes — you still see every model on the site.
- If you're on BASIC: you now have access to 5 additional models (WR10, PHANTOM, APEX, STREAK, ORACLE) that used to require PRO.
- If you're on FREE: you now get MIND and OPEN, two models with distinct strategies, immediately.
This is the third tier rebalance we've done this year, and it won't be the last. Models get re-evaluated quarterly — if VANGUARD slips below 70% next quarter, it moves down too. Same rule for everyone.
See the full breakdown on the Models leaderboard or the predictions page.