Node HandbookNODE HANDBOOK中文
Verified 2026-11-22

Free v2ray nodes, verified daily

30 node configurations, each requested from two independent exit points before it was listed. Measured latency and throughput are shown, including the ones that look bad.

Last rebuilt . This page is regenerated from the daily verification run; the figures below are that run's output, not a template.

Today's verified list

30 nodes across 16 regions and 4 protocols: HTTP, SS, VLESS, VMESS. Every row passed ten HTTPS requests from each of the two exit points; the request column shows both tallies, so a node with 9/10 on one side is visibly weaker rather than silently dropped.

NodeExitProtocolRequests (exit A / B)Latency medianFixed download / throughputConfig
node-023JPVLESS10/10, 10/100.380s / 0.419s256 KB, both succeeded · 367366 / 313944 B/sYAML
node-027JPVLESS10/10, 10/100.360s / 0.442s256 KB, both succeeded · 370977 / 297890 B/sYAML
node-030SGSS10/10, 10/100.110s / 0.535s256 KB, both succeeded · 1090984 / 215934 B/sYAML
node-057SGSS10/10, 10/100.122s / 0.549s256 KB, both succeeded · 420706 / 201804 B/sYAML
node-002USVMESS10/10, 10/100.595s / 0.046s256 KB, both succeeded · 187108 / 2647919 B/sYAML
node-006USSS10/10, 10/100.631s / 0.061s256 KB, both succeeded · 148733 / 1440351 B/sYAML
node-085SGVLESS10/10, 10/100.180s / 0.706s256 KB, both succeeded · 815804 / 185129 B/sYAML
node-047DEVLESS10/10, 9/100.721s / 0.576s256 KB, both succeeded · 179395 / 189684 B/sYAML
node-081FRVLESS10/10, 10/100.764s / 0.668s256 KB, both succeeded · 40571 / 36073 B/sYAML
node-003USVLESS10/10, 10/100.783s / 0.050s256 KB, both succeeded · 170000 / 3640888 B/sYAML
node-083TWVMESS10/10, 10/100.495s / 0.790s256 KB, both succeeded · 370269 / 183061 B/sYAML
node-045DEVLESS10/10, 10/100.833s / 0.531s256 KB, both succeeded · 126889 / 211919 B/sYAML
node-039NLVLESS10/10, 10/100.837s / 0.567s256 KB, both succeeded · 134241 / 210219 B/sYAML
node-007USVLESS10/10, 10/100.848s / 0.082s256 KB, both succeeded · 157868 / 1833174 B/sYAML
node-066NLVLESS10/10, 10/100.855s / 0.654s256 KB, both succeeded · 32309 / 99409 B/sYAML
node-034DESS10/10, 10/100.911s / 0.820s256 KB, both succeeded · 148466 / 176053 B/sYAML
node-028USVLESS10/10, 10/100.928s / 0.647s256 KB, both succeeded · 128365 / 212606 B/sYAML
node-035ATVMESS10/10, 10/100.953s / 0.702s256 KB, both succeeded · 78456 / 124533 B/sYAML
node-032UAVLESS10/10, 10/100.972s / 0.875s256 KB, both succeeded · 66188 / 161220 B/sYAML
node-064DEVLESS10/10, 10/100.975s / 0.674s256 KB, both succeeded · 45455 / 201030 B/sYAML
node-036DKVLESS10/10, 10/100.997s / 0.675s256 KB, both succeeded · 125324 / 23822 B/sYAML
node-001USVLESS10/10, 10/101.023s / 0.280s256 KB, both succeeded · 29834 / 102761 B/sYAML
node-026PLVLESS10/10, 10/101.026s / 0.637s256 KB, both succeeded · 133347 / 201649 B/sYAML
node-042AEVLESS10/10, 10/100.613s / 1.122s256 KB, both succeeded · 24006 / 18993 B/sYAML
node-038LVVLESS10/10, 10/101.144s / 0.888s256 KB, both succeeded · 127128 / 171224 B/sYAML
node-052USVLESS10/10, 10/101.091s / 1.146s256 KB, both succeeded · 144746 / 137248 B/sYAML
node-037KRVLESS10/10, 10/101.147s / 0.646s256 KB, both succeeded · 119202 / 223101 B/sYAML
node-084VNVLESS10/10, 10/100.295s / 1.168s256 KB, both succeeded · 539517 / 119809 B/sYAML
node-033SEVLESS10/10, 10/101.221s / 0.968s256 KB, both succeeded · 103778 / 177484 B/sYAML
node-093USHTTP10/10, 10/101.235s / 0.585s256 KB, both succeeded · 124846 / 151703 B/sYAML

By region

Exit regionNodesProtocols
US8HTTP, SS, VLESS, VMESS
DE4SS, VLESS
SG3SS, VLESS
JP2VLESS
NL2VLESS
AE1VLESS
AT1VMESS
DK1VLESS
FR1VLESS
KR1VLESS
LV1VLESS
PL1VLESS
SE1VLESS
TW1VMESS
UA1VLESS
VN1VLESS

What the numbers mean

Requests (exit A / B) is the count of successful HTTPS requests out of ten, made separately from a local network and a VPS. Two columns exist because one exit agreeing with itself proves little.

Latency median is the middle value of those requests, not the best one. A node with a single fast sample and eight slow ones shows the slow figure.

Fixed download is a 256 KB transfer, and the throughput beside it is the measured rate on each side. Throughput and latency disagree often enough that ranking on latency alone would mislead you.

Why a working node stops working

Free endpoints change without notice, and the failure is usually one of four: the subscription URL now returns an HTML page instead of a list, the body no longer parses, the address resolves but the port is closed, or the node is reachable and the exit is blocked. Telling those four apart takes a couple of commands and decides whether re-importing can help at all.

If the client reports success and nothing loads, the fault is downstream of the node: system proxy and TUN capture different traffic, and name resolution can fail while the tunnel is up.

Using these configurations

Each row links to a single-node YAML. Download it, then import it as a local profile: Clash Verge Rev, v2rayN, v2rayNG, Hiddify or sing-box. If what you have is a share link rather than a file, the subscription converter turns vmess, ss, trojan and vless links into a Clash configuration in the browser.

Nothing here is a paid service and nothing is promised to stay online. These are public endpoints that happened to pass a measurement today. Treat the exit as untrusted: do not log in to anything that matters through a node whose operator you cannot identify.

If you arrived looking for a free v2ray config to paste into one field rather than a file, take the YAML link on the row you want and open it: it is a single complete Clash configuration, not a fragment, so it can be imported without editing. If you have a share link instead of a file, the converter on that page turns vmess, vless, trojan and ss links into the same shape.

Questions

Are these nodes free to use?

They are configurations collected from public sources and re-tested here. Being reachable is not the same as being safe: you do not know who operates the exit, so do not send credentials or payment details through them.

How were the nodes tested?

Each one was used from two independent exit points, ten HTTPS requests from each, then a fixed-size download to measure throughput. A node only appears here if both sides completed the requests.

Why does a node that passed yesterday fail today?

Because the operator changed something, the IP was blocked, or the port closed. Free endpoints are not a service with an obligation to stay up. Re-fetch the subscription before assuming your client broke.

How do I import the YAML?

In Clash Verge Rev, open Profiles, then Import from file. In v2rayN the equivalent is Servers, then Import bulk URL from clipboard. Both are covered step by step in the client setup guides.

Is this a free v2ray subscription?

No. A subscription is a single URL the client re-fetches on a schedule; what you see here is a list of individual free v2ray configs, each re-tested that morning. If your client only accepts a subscription URL, download the daily bundle and serve it yourself, or rebuild the list from the rows - every node here is a complete standalone YAML.

How often is this list updated?

Every time a measurement batch is published, which is roughly once a day. The date at the top of the page is the batch the rows came from, and the free v2ray nodes daily update keeps the passing rows while dropping the ones that stopped answering. A node that disappears from the table did not pass the last run, so the absence is the signal.

All English guides · Daily archive (Chinese) · Error string index