Results
One report per reward epoch, published when Flare settles it, wins and losses in the same format, including what each epoch actually returned to delegators and stakers. Every number comes from the network's official settlement data (flare-foundation/fsp-rewards), never from our own claims. Charts and the machine-readable history table are below; live entity status is on the status page.
Where we stand point-in-time
Nothing on this page is typed in by hand. The two left tiles are
point-in-time values the publish pipeline derives at
generation time, active stake read live from Flare's PChainStakeMirror
contract, feed count from our machine-readable control plane, and stamps
with the moment they were read. The two right tiles, and everything below,
are settled per-epoch data from epochs.json.
The pipeline regenerates the whole file each epoch, so this page updates
itself as epochs settle; each tile's exact source is in its
data-source attribute and in epochs.json itself.
A new report publishes automatically once Flare sets the epoch's reward
Merkle root on chain, the pipeline checks every two hours, and won't
publish from files the chain hasn't confirmed yet.
At a glance settled data
Charted from the same machine-readable epochs.json anyone can fetch. Full history table below · live entity status: status →
FTSO accuracy per epoch settled data
Share of votes within 0.5% of the consensus median, per reward epoch, with the numerator and denominator in each point's tooltip. The dashed line is the 80% FIP.10 reward gate. Full fractions are in the table below.
Realized APR per epoch settled data
Realized historical rates from each epoch's actual payout file, reward ÷ weight, annualized simply over 104.36 epochs/yr, no compounding, not a projection. A zeroed epoch really pays 0.00%. Tooltips include that epoch's network medians for comparison.
Epoch history settled data
| Epoch | Delegation | Staking | Verdict | Dates | FTSO | FDC | Uptime* | Fast updates |
|---|---|---|---|---|---|---|---|---|
| Loading epochs.json… | ||||||||
*Uptime is a point-in-time read of the validator at settlement, not an epoch average, the official record only attests the ≥80% pass/fail. Every other column is Flare's settled data, shown with numerator / denominator.
curl https://cashlab.network/epochs.json, epochs
ascending, every gate with numerator and denominator, plus links to the official
source files and the chain-verified merkle root per epoch.
Epoch 432Sep 10 – Sep 14, 2026 · ended Sep 14PAID
All four gates passed, zero strikes. Pass banked — buffer now 3.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
99.35% 206,961 / 208,320 · needs 80% · rank 59 of 100 providers · network median 99.53% · 98 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
100.0% 3,359 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9977% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
1263 vs 972 expected · needs ≥972 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
1.92% APR · network median 1.35% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
9.62% APR · network median 6.77% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year — no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 91 providers that paid out in
this epoch. 41,882.43 FLR went to delegators and stakers this epoch; our fee was 10,470.61 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Epoch 431Sep 7 – Sep 10, 2026 · ended Sep 10PAID
All four gates passed, zero strikes. Pass banked — buffer now 3.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
99.59% 207,470 / 208,320 · needs 80% · rank 43 of 100 providers · network median 99.42% · 97 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
99.9% 3,356 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9975% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
1236 vs 964 expected · needs ≥964 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
1.91% APR · network median 1.37% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
9.54% APR · network median 6.87% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year — no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 92 providers that paid out in
this epoch. 41,535.34 FLR went to delegators and stakers this epoch; our fee was 10,383.83 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Epoch 430Sep 3 – Sep 7, 2026 · ended Sep 7PAID
All four gates passed, zero strikes. Pass banked — buffer now 3.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
99.43% 207,124 / 208,320 · needs 80% · rank 67 of 100 providers · network median 99.62% · 97 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
100.0% 3,360 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9972% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
880 vs 631 expected · needs ≥631 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
2.18% APR · network median 1.38% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
10.90% APR · network median 6.89% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year — no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 94 providers that paid out in
this epoch. 26,952.35 FLR went to delegators and stakers this epoch; our fee was 6,738.09 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Epoch 429Aug 31 – Sep 3, 2026 · ended Sep 3PAID
All four gates passed, zero strikes. Pass banked — buffer now 3.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
99.12% 206,497 / 208,320 · needs 80% · rank 79 of 100 providers · network median 99.72% · 98 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
100.0% 3,359 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9969% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
887 vs 638 expected · needs ≥638 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
2.11% APR · network median 1.37% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
10.57% APR · network median 6.83% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year — no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 96 providers that paid out in
this epoch. 26,140.40 FLR went to delegators and stakers this epoch; our fee was 6,535.10 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Epoch 428Aug 27, Aug 31, 2026 · ended Aug 31PAID
All four gates passed, zero strikes. Pass banked, buffer now 3.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
98.97% 206,169 / 208,320 · needs 80% · rank 87 of 100 providers · network median 99.75% · 97 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
99.9% 3,356 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9964% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
649 vs 480 expected · needs ≥480 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
2.41% APR · network median 1.37% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
12.06% APR · network median 6.84% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year, no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 94 providers that paid out in
this epoch. 20,460.47 FLR went to delegators and stakers this epoch; our fee was 5,115.12 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Epoch 427Aug 24, Aug 27, 2026 · ended Aug 27PAID
All four gates passed, zero strikes. Pass banked, buffer now 3.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
98.54% 205,278 / 208,320 · needs 80% · rank 86 of 100 providers · network median 99.79% · 98 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
100.0% 3,359 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9961% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
704 vs 491 expected · needs ≥491 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
2.41% APR · network median 1.41% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
12.05% APR · network median 7.04% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year, no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 91 providers that paid out in
this epoch. 20,446.50 FLR went to delegators and stakers this epoch; our fee was 5,111.63 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Epoch 426Aug 20, Aug 24, 2026 · ended Aug 24PAID
All four gates passed, zero strikes. Pass banked, buffer now 2.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
97.01% 202,082 / 208,320 · needs 80% · rank 94 of 100 providers · network median 99.71% · 98 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
99.3% 3,337 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9952% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
691 vs 496 expected · needs ≥496 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
2.48% APR · network median 1.42% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
12.38% APR · network median 7.11% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year, no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 93 providers that paid out in
this epoch. 21,005.10 FLR went to delegators and stakers this epoch; our fee was 5,251.27 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Epoch 425Aug 17, Aug 20, 2026 · ended Aug 20PAID
All four gates passed, zero strikes. Pass banked, buffer now 1.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
96.36% 200,744 / 208,320 · needs 80% · rank 95 of 100 providers · network median 99.79% · 99 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
99.7% 3,351 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9944% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
215 vs 133 expected · needs ≥133 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
4.07% APR · network median 1.50% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
20.34% APR · network median 7.48% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year, no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 96 providers that paid out in
this epoch. 5,949.66 FLR went to delegators and stakers this epoch; our fee was 1,487.42 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Epoch 424Aug 13, Aug 17, 2026 · ended Aug 17PAID
All four gates passed, zero strikes.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
95.36% 198,655 / 208,320 · needs 80% · rank 94 of 100 providers · network median 99.82% · 98 of 100 cleared the floor | ✓ PASS |
| FDC participation attestation rounds rewarded |
99.1% 3,331 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9998% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
196 vs 131 expected · needs ≥131 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
4.27% APR · network median 1.56% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
21.37% APR · network median 7.79% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year, no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 95 providers that paid out in
this epoch. 6,250.37 FLR went to delegators and stakers this epoch; our fee was 1,562.59 FLR.
Our method reproduces Flare's own published per-provider rates exactly.
Scrutiny cuts both ways: delegation reward per unit of registration weight trends higher for smaller providers (r ≈ −0.25, epoch 424, our derivation; the trend reverses for total rewards including staking), and CashLab is the 4th-smallest of 100 registered providers by registration weight — so part of our delegation APR edge is size, not skill. Derived 2026-08-19 from the same public settlement files linked below.
Epoch 423Aug 10, Aug 13, 2026 · ended Aug 13PAID
Our first paying epoch. All four gates passed, zero strikes.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
94.03% 199,043 / 211,680 · needs 80% | ✓ PASS |
| FDC participation attestation rounds rewarded |
99.4% 3,340 / 3,360 · needs 60% | ✓ PASS |
| Validator uptime 3,000,000 FLR self-bond |
99.9997% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
197 vs 141 expected · needs ≥141 | ✓ PASS |
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
4.35% APR · network median 1.65% | PAID |
| Staking (mirror) what a staker earned, net of our 20% fee |
21.75% APR · network median 8.24% | PAID |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year, no compounding assumed. The network median covers the 91 providers
that paid out in this epoch. 6,361.67 FLR went to delegators and stakers this epoch;
our fee was 1,590.42 FLR.
What changed between 422 and 423
Everything in 422's "what changed" list ran through this epoch, our own node as primary RPC, proven fallbacks on every critical source, work-based health checks, off-site alerting. The result: FDC recovered from 35.6% to 99.4% (missing only 20 of 3,360 rounds), and FTSO accuracy cleared the gate with 14 points of margin. Two new feeds (SGB/USD and USDX/USD) shipped mid-epoch and hit the accuracy band in essentially every round they were live for; epoch 424 is their first full epoch.
Epoch 422Aug 6, Aug 10, 2026 · ended Aug 10PAID ZERO
Our first reward-eligible epoch, and it paid nothing. Here is exactly what happened.
The four gates
| FTSO accuracy votes within 0.5% of consensus median |
74.85% 158,438 / 211,680 · needs 80% | ✗ FAIL |
| FDC participation attestation rounds rewarded |
35.6% 1,197 / 3,360 · needs 60% | ✗ FAIL |
| Validator uptime 3,000,000 FLR self-bond |
99.9995% needs 80% | ✓ PASS |
| Fast updates block-latency submissions |
151 vs 136 expected · needs ≥136 | ✓ PASS |
Two gates failed with no passes banked, under FIP.10 each failed condition needs a banked pass to absorb it, so the epoch zeroed entirely, including the staking side, despite the validator running clean throughout.
What it returned
| Delegation (wNat) what a delegator earned, net of our 20% fee |
0.00% APR · network median 1.54% | NIL |
| Staking (mirror) what a staker earned, net of our 20% fee |
0.00% APR · network median 7.69% | NIL |
Rates are realized, not projected: derived from the epoch's actual payout file
(reward-distribution-data.json) as reward ÷ weight, annualized simply over
104.36 epochs per year, no compounding assumed, because rewards must be claimed and
re-delegated to compound. The network median covers the 90 providers that paid out in
this epoch. Our method reproduces Flare's own published per-provider rates exactly.
What happened
Our RPC layer had a single point of failure. A metered third-party endpoint we depended on was disabled mid-epoch over a billing overage, our oversight, causing 14 consecutive hours of zero submissions (Aug 9, 07:00-20:00 UTC). Outside that window our FTSO hit rate was ~89.8%; the outage alone dragged it under the gate. Separately, a health check that trusted "process running" over "work coming out" hid a dead Ethereum data source for roughly two days, which is what pushed FDC participation down to 35.6%.
What changed since, all live and tested
- Our own Flare node is now the primary RPC for every component, with automatic failover through a monitored pool, a dead endpoint rotates out within ~6 minutes.
- Every critical data source now has a proven fallback, including the XRP feed (89% of FDC rounds), validated by live drill, not by assumption.
- Health checks now require proven work (a real verified transaction, a climbing submission counter), not just a running process.
- Independent off-site monitoring alerts us even if our servers die entirely.
- Feed-accuracy improvements measured and merged (+1.44pp), with more in test.