← Back to archive

Published: February 21, 2026

Daily AI usage report (2026-02-21)

Measured window: 2026-02-20T08:12:04.397953+00:00 → 2026-02-21T08:12:04.397953+00:00 (rolling 24h).

Totals: Input 995,559, Output 65,342, Total 1,060,901 tokens.

Estimated MiniMax cost: M2.5 $0.3771 · M2.5-highspeed $0.7542.

Math

Input cost (M2.5) = 995,559/1,000,000 × 0.3 = 0.298668 USD
Output cost (M2.5) = 65,342/1,000,000 × 1.2 = 0.078410 USD
Total M2.5 = 0.377078 USD

Input cost (M2.5-highspeed) = 995,559/1,000,000 × 0.6 = 0.597335 USD
Output cost (M2.5-highspeed) = 65,342/1,000,000 × 2.4 = 0.156821 USD
Total M2.5-highspeed = 0.754156 USD

Visualization

Input tokens Output tokens Cost scale

Rolling efficiency score

Score: 42.7 / 100 (Needs optimization) over 7 recent snapshot window(s).

Window totals: Input 4,764,812, Output 278,472, Total 5,043,284 tokens, M2.5 est. cost $1.7636.
Output share: 5.52% · Cost per 1k tokens: $0.0003

Optimization guidance

If score drifts down, route more low-risk transforms to cheaper/local models, tighten premium escalation triggers, and reduce verbose multi-step loops on routine tasks.

Notes

This report uses observed model usage records from session transcripts over a rolling 24-hour window.