13 dollas'
ANTIGRAVITY PRO USAGE REPORT
Date: 2026-11-01
Plan: Google AI Pro ($20/mo, g1-pro-tier), user-confirmed
Model: Gemini 4 Argon, enum 1417, backend string gemini-4-argon-e
Source: 4 local conversation DBs, gen_metadata only (no prompts/code read)
User report: 2 sessions used 64% of weekly quota
1. EXACT TOKEN COUNT
------------------------------------------------------------------------------
conversation gens uncached in cache reads output total
1 28 265,395 2,296,626 36,916 2,598,937
2 60 544,569 7,008,431 65,370 7,618,370
3 34 232,062 4,487,715 82,759 4,802,536
4 66 650,786 8,260,582 114,357 9,025,725
TOTAL 188 1,692,812 22,053,354 299,402 24,045,568
Output includes 171,632 thinking tokens.
Cache reads = 91.7% of all tokens, 92.9% of prompt tokens.
Blended usage = 24,045,568 tokens.
2. MAIN CALC - CURRENT API PRICES AS OF 2026-11-01
------------------------------------------------------------------------------
Gemini 4 Argon intro rates, valid through 2026-11-30:
input $2.00/M, output $10.00/M, cache read $0.10/M
component tokens rate cost
uncached in 1,692,812 $2.00/M $3.3856
cache reads 22,053,354 $0.10/M $2.2053
output 299,402 $10.00/M $2.9940
TOTAL 24,045,568 $8.5850 -> $8.58
Blended API-equivalent price: $0.357 per 1M tokens.
Secondary price, from 2026-12-01 ($4/$20/$0.20):
same tokens = $17.17.
3. QUOTA MATH FOR PRO
------------------------------------------------------------------------------
64% of weekly = 24,045,568 tokens
Full week at same mix = 24,045,568 / 0.64 = 37,571,200 tokens
If Pro meters Argon at its listed API price:
64% = $8.58, implied full weekly bucket = $13.41