filecoin network intelligence

Powered by Lantern

The real state of the network, not the dashboard version.

Snapshot 2026-06-18 18:32 UTC, mainnet chain epoch 6116703. 622 reported miner IDs collapse into 267 deduplicated operators. QA / raw ratio is 8.8×, so most of the network is verified-deal sectors, not committed capacity.

622
Reported miner IDs
267
Real operators
1.7EiB
Raw storage
15.0EiB
QAP (8.8× inflated)
83
Run 90% of power
7
Run 50% of power
60M+
Active deals (est.)
70M+FIL
Pledge collateral

Power concentration

83 entities run 90% of Filecoin's storage. 7 entities run half. The 622 reported miner IDs deduplicate into 267 unique operators by shared owner / worker / control / beneficiary addresses.

Is the network growing or shrinking?

24 network meter
Declining -125.36 PiB

Net change in raw network power across all active miners during the most recent measurement period. The network is shedding storage compared with the prior reading.

Collapsing Declining Stable Growing Booming

Network power history

0 EiB6 EiB12 EiB19 EiB25 EiBOct 20Jul 21Apr 22Jan 23Oct 23Jul 24Apr 25Jan 26Jun 26peak 16.77 EiB · Jul 22QAP peak 25.0 EiB · Jan 24Raw power (EiB)Quality-adjusted (EiB)
Quarterly snapshots from chain history; latest point is today's snapshot. Raw peaked at 16.77 EiB (Jul 2022); QAP peaked at ~25 EiB (Dec 2023). Today the QA/raw ratio is 8.8×.

Operators required to reach a power threshold

7
For 50%
29
For 75%
83
For 90%
120
For 95%
190
For 99%

How the top tiers stack up

Top 1: 12%
Ranks 2-5: 31%
Ranks 6-10: 17%
Ranks 11-20: 11%
All others: 30%

Verified deals vs committed capacity

Deal: 1508 PiB (86.4%)
CC: 238 PiB

Cumulative power curve: reading the concentration

Operators are sorted from largest to smallest along the X axis. The curve plots the cumulative share of network power once you've added that many operators. The faster the line climbs, the more concentrated the network. A perfectly equal network would be a straight diagonal; a single-operator network would be a vertical line at the left edge.

0% 25% 50% 75% 100% % of network power Top operator (rank 1) Operator rank → All 267 operators 50% threshold, reached at operator #7 90% threshold, reached at operator #83
Cumulative share of network raw power across 267 operators Top 7 of 267 operators hold half the network

Key findings

Real raw storage is 1.71 EiB. The QAP metric (15.0 EiB) hides the picture behind a 10× verified-deal multiplier. Today the QA / raw ratio is 8.8×.

86.4% of remaining storage is verified-deal data. Only 238 PiB (13.6%) is committed capacity (CC).

Top 30 operators

Ranked by raw byte power. 7 of these control half the network.

# Owner Miner IDs Raw power Share Region Reach Sample miner IDs
1 f02219535 +4 more 40 201.8 PiB 11.73% 6/40 f01081394, f01081419 +38
2 f02950326 +7 more 38 195.9 PiB 11.39% 4/38 f01859603, f01923786 +36
3 f02829362 +3 more 11 121.7 PiB 7.08% 1/11 f01106668, f01315096 +9
4 f01986938 +8 more 43 118.4 PiB 6.88% 2/43 f03099777, f03100014 +41
5 f02366305 2 88.7 PiB 5.16% 0/2 f0377277, f0407287
6 f02090080 +16 more 23 74.2 PiB 4.31% 0/23 f02825281, f02825675 +21
7 f03650877 15 64.7 PiB 3.76% 0/15 f02221110, f02831202 +13
8 f01620757 +6 more 15 58.7 PiB 3.41% 8/15 f01111110, f01234567 +13
9 f01216600 +6 more 10 56.7 PiB 3.30% 8/10 f01989013, f01989014 +8
10 f01438778, f01811564 7 31.4 PiB 1.82% 6/7 f01422327, f02252023 +5
11 f02951814 +4 more 10 31.1 PiB 1.81% 0/10 f02211572, f02948413 +8
12 f02051046 +4 more 16 30.5 PiB 1.77% 0/16 f01882035, f01939377 +14
13 f03200943, f03278453 10 30.0 PiB 1.74% 1/10 f03275917, f03276663 +8
14 f03644272 +4 more 5 15.5 PiB 0.90% 0/5 f03644142, f03645007 +3
15 f03632590 4 14.6 PiB 0.85% 4/4 f03623016, f03623017 +2
16 f01662596 +21 more 22 14.4 PiB 0.83% 0/22 f0101020, f0101021 +20
17 f02247755 1 13.6 PiB 0.79% 0/1 f01780352
18 f03769068, f03769069 2 13.5 PiB 0.79% 0/2 f03100008, f03100009
19 f03085528, f03669997 6 13.5 PiB 0.78% 0/6 f03252730, f03378086 +4
20 f02088713 5 13.3 PiB 0.77% 2/5 f03231154, f03312989 +3
21 f03066523 2 12.9 PiB 0.75% 2/2 f01697248, f01904546
22 f03742980 3 11.7 PiB 0.68% 0/3 f02808877, f03226666 +1
23 f03053508, f03233419 5 11.4 PiB 0.67% 0/5 f01313, f01708981 +3
24 f02951819 5 11.1 PiB 0.64% 0/5 f03066557, f03066866 +3
25 f03644092, f03653941 5 10.8 PiB 0.63% 0/5 f03145504, f03192503 +3
26 f03543078 5 10.1 PiB 0.58% 3/5 f03372222, f03373333 +3
27 f01916370 1 10.0 PiB 0.58% 0/1 f01083914
28 f01264531 +4 more 5 8.7 PiB 0.51% 0/5 f01265268, f01546926 +3
29 f03524241, f03524244 2 8.6 PiB 0.50% 1/2 f03499999, f03500000
30 f03728578, f03728581 2 8.1 PiB 0.47% 0/2 f0109903, f03498626

Storage providers exiting the network

83 miners are currently losing power (net -3.15 PiB raw across all decliners this measurement period). When power drops to zero the SP is effectively exiting; pledge collateral unlocks back into circulation.

# Miner ID Region Current power Change Decline %
1 f02243688 4.50 PiB -1.360 PiB -23.2%
2 f0678845 7.87 PiB -0.549 PiB -6.5%
3 f03099981 4.35 PiB -0.149 PiB -3.3%
4 f0407287 30.58 PiB -0.140 PiB -0.5%
5 f0377277 58.16 PiB -0.140 PiB -0.2%
6 f03528979 0.78 PiB -0.122 PiB -13.5%
7 f01755999 4.88 PiB -0.117 PiB -2.3%
8 f01084941 7.38 PiB -0.100 PiB -1.3%
9 f01546926 0.38 PiB -0.075 PiB -16.4%
10 f0866468 0.82 PiB -0.072 PiB -8.0%
11 f0861587 1.54 PiB -0.065 PiB -4.0%
12 f0872282 5.32 PiB -0.061 PiB -1.1%
13 f0187709 0.57 PiB -0.042 PiB -6.9%
14 f03112580 1.54 PiB -0.026 PiB -1.7%
15 f01431043 0.38 PiB -0.022 PiB -5.6%
16 f03242023 0.09 PiB -0.016 PiB -15.2%
17 f01019009 0.46 PiB -0.012 PiB -2.6%
18 f0866680 0.45 PiB -0.012 PiB -2.5%
19 f01996719 3.47 PiB -0.010 PiB -0.3%
20 f01016847 0.45 PiB -0.010 PiB -2.1%

Storage software

Daemon detected per miner ID via libp2p identify. private = peer ID published on chain but not dialable; no peer ID published = no libp2p ID at all on chain; custom software = dialed and got an unrecognised agent string (could be a fork / private build).

private (firewalled)
488
78.5%
no peer ID published
60
9.6%
boost
51
8.2%
curio
21
3.4%
droplet
2
0.3%

Chain nodes

Public-facing Lotus, Forest, and Venus full nodes our daemon is currently connected to via libp2p. Lotus and lotus-miner are different things: Lotus is the chain consensus daemon, lotus-miner is the (deprecated) SP daemon.

lotus
125
91.9%
forest
5
3.7%
boost
3
2.2%
venus
2
1.5%
custom software
1
0.7%

FoC providers

PDP services registered on the on-chain ServiceProviderRegistry that are actually serving traffic. We only list HTTP-reachable providers.

#NameService URLRegionHTTP
13 Angelo https://curio.akave.xyz US 200
6 Axiaoming_ZHU https://zpdp.oplian.com HK 200
2 beck-main https://pdp-main.660688.xyz:8443 CN 200
1 ezpdpz-main https://main.ezpdpz.net GB 200
30 foc-mainnet https://foc.dealbot.site CA 200
7 infrafolio-mainnet-pdp https://mainnet-pdp.infrafolio.com US 200
5 Mongo2Stor Mainnet https://pdp.lotus.dedyn.io US 200
16 pdp main filstarry https://pdp2.filstarry.com HK 200
17 pg-pdp-mainnet https://pdp-mainnet.pixelguardian.eu LT 200
9 ruka-main https://ruka.drongyl.com HK 200
15 TopBlocks https://pdp-main.topblocks.io US 200
29 TSE REIERSEN https://sp.reiers.io NO 200

Supply context

FIL flows backing the storage layer. Pledge collateral is what SPs lock to commit storage; market collateral is what providers lock per active deal.

78,030,695FIL
SP pledge collateral
519,929FIL
Provider deal collateral
124,795,612
DataCap allocations
749782
Miners ever registered
631
Active right now

Methodology

How we counted

1. Source data. Filecoin chain state via our own Lotus mainnet node JSON-RPC. Active SP set from Filfox /miner/list/power. FoC providers from on-chain ServiceProviderRegistry contract 0xf55dDbf63F1b55c3F1D4FA7e339a68AB7b64A5eB.

2. Operator deduplication. Union-find clustering on shared Owner, Worker, Control, Beneficiary addresses, plus shared IPs (with a 50-miner CDN cap). Zero-power miner IDs are excluded.

3. Software detection. libp2p identify probe; agent string + protocol set classified into Curio / Boost / lotus-miner / Venus-miner / Droplet / private / no-peer-id / custom software.

4. Power split. QAP = raw + 9 × raw_verified, so verified-raw ≈ (QA − raw) / 9. CC = raw − verified.

5. IP geolocation. db-ip.com free GeoLite2-compatible mmdb (city + ASN, monthly).

Limitations. Operator dedup is structural-only. SP dial reachability is a single-vantage probe. GeoIP reflects announced multiaddr IPs; VPN / relay endpoints can mislead.