Reward Examples

This page does the arithmetic the chain does when a vote lands and when a post pays out, with every intermediate figure, so that a reader can check a payout in their own history or predict one. The rules are explained on Proof-of-Brain, Voting and Curation and Posting and Rewards; the constants are on Chain Parameters. Here there are only numbers, integer arithmetic throughout, as in the code. Amounts are in base units where it matters: 1 PIXA = 1,000 units, 1 VESTS = 1,000,000 units.

The formulas

Python# (a) a vote -> rshares                          hive_evaluator_social.cpp:660-718 (v1.30.0)
used_mana = ceil( E * |weight| * 86_400 / 10_000 / 4_320_000 )   # E: effective Pixa Power in VESTS units
rshares   = max(0, used_mana - 50_000)                           # 50,000 since HF30; 50,000,000 before
if seconds_to_payout < 43_200: rshares = rshares * seconds_to_payout // 43_200

# (b) a post's claim on the fund                 util/reward.cpp:7-32, s = 2,500 (content constant)
claim = ((r + s)**2 - s**2) // (r + 4*s)        # r: the post's net rshares; about r - 5,000 for large r

# (c) the fund's denominator, every block        database_comment.cpp:409
recent_claims -= recent_claims * 3 // 1_296_000   # decays by e^(-t / 15 days); every paying post adds its claim

# (d) the payout                                 util/reward.cpp:17-29
payout = reward_balance * claim // (recent_claims + claims paying out in this block)
if payout in PXS at the median feed < 0.020 PXS: payout = 0

# (e) the split                                  database_comment.cpp:227-228, 281-301; database.cpp:979-1035 (v1.30.0)
curation = payout * 40 // 100;  author = payout - curation
liquid_half  = author // 2;  vested_half = author - liquid_half
author_pxs   = liquid_half * print_rate // 10_000, converted to PXS at the median feed, rounded down to 0.001 PXS
author_pixa  = liquid_half - ceil(author_pxs in PIXA)   # since HF30: the PIXA that did not fit a whole 0.001 PXS is paid, not burned
author_vests = vested_half, converted at the VESTS price

# (f) a vote's curation weight                   hive_evaluator_social.cpp:784-831
W(x) = x // approx_sqrt(x + 2*s)                # approx_sqrt: the chain's integer approximation
weight = W(rshares before + this vote) - W(rshares before)
weight //= 1 in the first 24 h, 2 until 72 h, 8 after
curator's share = curation * weight // sum of weights

The late-vote factor in (a), the age factors in (f) and the dust test in (d) are the three places where timing changes an amount. Everything else is proportion.

Example 1: one full vote from 10,000 Pixa Power

An account with E = 10,000 Pixa Power (10,000,000,000 VESTS units), full mana, votes 100% on a fresh post with no other votes.

(a) Mana and rshares. used_mana = ceil(10,000,000,000 × 10,000 × 86,400 / 10,000 / 4,320,000) = 200,000,000, exactly 2% of the bar. rshares = 200,000,000 − 50,000 = 199,950,000.

The threshold. A vote produces rshares when its mana exceeds 50,000, which for a full vote means more than 2.5 Pixa Power: E = 2.500000 VESTS gives 0, E = 2.500001 VESTS gives 1. Before hardfork 30 the deduction was 50,000,000 and the line was 2,500 Pixa Power: a 100-Pixa-Power account's full vote spent 2,000,000 mana and produced nothing; today it produces 1,950,000 rshares (Voting and Curation).

(b) to (d) Payout, with the fund as read on 2026-10-08: reward_balance 517,193.258 PIXA, recent_claims 9,930,574,867,540, median feed 1.000 PXS = 51.833 PIXA (a placeholder; PXS promises no price).

claim  = ((199,950,000 + 2,500)² − 2,500²) / (199,950,000 + 10,000) = 199,945,000
payout = floor(517,193,258 × 199,945,000 / (9,930,574,867,540 + 199,945,000)) = 10,413 units = 10.413 PIXA = 0.200 PXS   (≥ 0.020 PXS: paid)

(e) The split. Curation 40%: 4.165 PIXA, to the only curator as 4.164999 VESTS. Author 60%: 6.248 PIXA; the liquid half, 3.124 PIXA, becomes 0.060 PXS (floor(3,124 × 1,000 / 51,833)), and since hardfork 30 the 0.014 PIXA that did not fit into a whole 0.001 PXS is paid as PIXA instead of being burned (3,124 − ceil(60 × 51,833 / 1,000) = 14 units); the vested half, 3.124 PIXA, becomes 3.123999 VESTS.

The same vote at 30% gives 59,950,000 rshares and 3.121 PIXA (0.060 PXS). Before hardfork 30 it gave 10,000,000 rshares and 0.520 PIXA, below the minimum, so nothing.

What a lone full vote is worth today, with the same fund, for other stakes:

In each row the author also receives the few thousandths of a PIXA that did not fit into a whole 0.001 PXS. The smallest lone full vote that clears the 0.020 PXS minimum today comes from about 998.3 Pixa Power (19,916,369 rshares, 1.037 PIXA). A smaller vote is not wasted: its rshares add to the post's total. Eleven full votes from 100-Pixa-Power accounts make 21,450,000 rshares and pay 1.116 PIXA (0.021 PXS), and the eleven share the curation; ten make 19,500,000 and fall just under the line, because each vote loses its own 50,000. Before hardfork 30 any number of such votes produced nothing.

Live checks. The account wang, with 50 VESTS of received delegation, voted 100% after hardfork 30; the chain's effective_comment_vote recorded 950,000 rshares, as the table says. And the first payouts after hardfork 30 show the conversion remainder: @saitam/far-away-staying-awake-1790785626144, paid in block 954,672, received 16.814 PXS, 0.033 PIXA and 871.553993 VESTS. The liquid half was 871,554 units: floor(871,554 × 1,000 / 51,833) = 16,814 PXS units, 871,554 − ceil(16,814 × 51,833 / 1,000) = 33 units of PIXA. Before hardfork 30 those 33 units were burned and hive_payout was always 0 while the print rate was 100%.

Example 2: a real payout, reproduced to the unit

The post @bryan/horus-portrait-1790597187843, published in block 690,312 with the creation time 2026-09-28 12:06:27 (the time of the block before it, which is what the chain records), paid out 7 days later in block 891,898, 2026-10-05 12:06:27. This was before hardfork 30, so the deduction is 50,000,000. Three votes, read from the effective_comment_vote operations (account_history_api.enum_virtual_ops):

Curation weights (f). All three votes fell in the first 24 hours, so no age factor. mathiew: W(16,934,615,385) − W(0) = 130,130. matias: W(27,880,549,478) − W(16,934,615,385) = 32,069. bryan: 0 rshares, weight 0. The chain recorded 130,130, 32,069 and a total of 162,199.

Claim (b). net rshares 27,880,549,478; claim = 27,880,544,478 (5,000 less).

The fund at the payout block. The fund is read every block, so it was reconstructed from a snapshot 1,493 blocks later (block 893,391: 481,746.405 PIXA and 11,598,498,893,914) by undoing the decay and the issuance between them, 653 units a block in that window: reward_balance 481,925.934 PIXA, recent_claims 11,610,772,368,671 before this post's claim, 11,638,652,913,149 with it; it was the only post paying out in that block.

Payout (d). floor(481,925,934 × 27,880,544,478 / 11,638,652,913,149) = 1,154,459 units = 1,154.459 PIXA, which is 22.272 PXS at the feed.

Split (e). Curation floor(0.40 × 1,154,459) = 461,783; author 692,676.

The comment_reward operation recorded payout 22.272 PXS, total_payout_value 13.362 PXS (the author's part in PXS) and curator_payout_value 8.909 PXS, all matching. The VESTS figures use the reward VESTS price (total_vesting_shares + pending rewarded shares) / (fund + pending rewarded PIXA), about 0.9999999928 VESTS per PIXA at the time.

Pending payout check. The app's "pending payout" figure follows the same arithmetic with the fund of the moment. For @retro/peach-knight-1791211388189 at block 895,530: net rshares 24,319,843,531, claim 24,319,838,531, fund 483,143.172 PIXA = 9,321.150 PXS, recent_claims 11,541,212,091,810: floor(claim × 9,321,150 / recent_claims) = 19.641 PXS, which is what the effective_comment_vote operation carried as pending_payout (the account-history plugin computes it with the fund of the moment).

Example 3: when the vote lands

Voter A has 100,000 Pixa Power and votes 100% one hour after publication; voter B has 50,000 Pixa Power and votes 100% at different ages. Hardfork 30 rules; the fund of 2026-10-08.

A: 1,999,950,000 rshares, curation weight W(1,999,950,000) − W(0) = 42,642. (The chain's approx_sqrt(1,999,955,000) is 46,900 where the true root is 44,720; the approximation is part of the rule.)

Three things to read off the table. B's vote is half the size of A's but, cast second, its raw weight is about a quarter of A's: the square-root curve rewards being first. The age factors then divide it by 2 or 8. And a vote in the last 12 hours loses rshares too, so it raises the post's payout less: 130.165 instead of 156.190 PIXA.

Live check of the age factors. On @karo/diving-in-clear-waters-1790621283370 (created 2026-09-28 18:48:03), six votes at ages 0.3 h, 40 h, 45 h, 56 h, 94 h and 124 h have raw weights 102,020, 60,041, 88,338, 93,234, 45 and 42; the chain recorded 102,020, 30,020, 44,169, 46,617, 5 and 5: ×1, ÷2, ÷2, ÷2, ÷8, ÷8.

The constants used

Sources

• Code, at tag v1.30.0 for the vote evaluator and at commit 48f75a2 for the unchanged rest: hive_evaluator_social.cpp:660-718 (mana, dust, late damping) and 784-831 (curation weight); util/reward.cpp:7-71 (claim, payout, curves, approx_sqrt); database_comment.cpp:129-325, 395-444 (the split and the fund); database.cpp:979-1035 at v1.30.0 (the author's PXS and the remainder); supplement_operations.hpp:25-73 (pending payout); approx_sqrt test vector in basic_tests.cpp:594-597.
• Live chain: condenser_api.get_reward_fund ["post"] and get_dynamic_global_properties at block 977,074 (2026-10-08 11:07 UTC) and at blocks 893,391, 895,071, 895,416 (2026-10-05); account_history_api.enum_virtual_ops for blocks 690,312 to 891,898 (the votes and the payout of Example 2) and for @karo/diving-in-clear-waters-1790621283370; the effective_comment_vote operation with its pending_payout at block 895,530; the effective_comment_vote of wang after block 949,330; account_history_api.enum_virtual_ops for the author_reward operations in blocks 950,000 to 979,000.
• The arithmetic was run in a script with integer operations only; every chain figure quoted matches its output to the unit.