GPT IQ Index
Compare GPT model intelligence, quota status, and reset probability for the next 48 hours.
Radar score ยท Radar Bench v1
Page updated: Nov 3, 2026, 08:00 AMRadar score = percentage-grade ร 1.5 (max 150). The 64-task set covers software engineering, spatial reasoning, computer vision, terminal tooling, systems engineering, and scientific computing.
Price is mean API-equivalent cost, not subscription charges. Duration uses known-sample means. Scores use each task's latest three valid grades.
Quota radar
Historical Pro 20x data ยท Awaiting an updatePro 200 ยท Only GPT-6 Astra
$1,463.00
Provided by site owner
Pro 200 ยท Only GPT-6.1 Sol
$1,014.00
Provisionally provided by site owner
Pro 200 ยท Only GPT-5.6 Sol
$1,919.83
Measured allowance
Pro 200 ยท Only GPT-5.6 Luna
$1,145.10
Measured allowance
These are historical allowance estimates for the old Pro 20x (Pro 200) plan; current data is pending. Dollar figures follow the label on each card. They are not model unit prices, subscription charges, live balances, or a comparison of cost for the same task and runtime.
See the official source for current Standard API prices per 1M tokens. API-equivalent prices do not represent subscription allowances or actual charges. Official source
Prediction Radar: 48-hour reset probability
Day 21 reset completed, confirmed by the stationmaster; no pending delivery window. Original calibrated probabilities and calculation time are retained.
This month + the previous six months
Reset counts
Grouped by Beijing time; each reset or card issuance is counted once.