Shadowrocket supports the six protocols covered here, but a protocol name alone is not a complete configuration. This guide is for users who already have service settings and need to identify the Type or verify an imported profile. First, match the protocol in the table; then check credentials, transport, and certificate settings. Finally, use Global Routing to distinguish connection issues from rule issues. The steps below use the iPhone / iPad interface; layouts may differ on other platforms.
Six Protocols and Their Required Settings
In Shadowrocket, the protocol type determines how the client establishes a session with the remote server; it does not determine which websites use that connection. That is controlled by Global Routing and rules in Config. The table below helps identify six common protocols. It is not a complete list of supported protocols or extensions, and it is not a speed ranking.
| Protocol | Transport characteristics | What to check | Common mistakes |
|---|---|---|---|
| Shadowsocks | An encrypted proxy protocol that can carry TCP traffic. UDP availability also depends on the server and network. | Address, Port, Method, Password, and any additional settings required by the server. | Copying the password but missing the Method, or enabling obfuscation settings that the server has not configured. |
| VMess | Authenticates with parameters such as a UUID and can be paired with different transports and TLS. | UUID, transport, and TLS. With WebSocket, also check Host and Path. | The UUID is correct, but the WebSocket Path or TLS settings do not match the server. |
| VLESS | The protocol itself does not encrypt data; connection security depends on the configured security layer. | UUID, transport, security layer, and its settings. Check Flow too if it appears in your service configuration. | Treating VLESS as VMess, or assuming a UUID means no security-layer settings are needed. |
| Trojan | Typically authenticates with a password and carries proxy traffic over a TLS connection. | Password, Address, Port, SNI, and certificate verification settings. | Entering a UUID as the Password, or using the connection IP as the certificate hostname. |
| WireGuard | An encrypted tunnel over UDP that uses public and private keys to establish a peer connection. | Local Private Key, tunnel address, peer Public Key, and Endpoint. Check settings such as MTU against your configuration. | Reversing the public and private keys, or omitting the tunnel address assigned by the server. |
| Hysteria2 | Built on QUIC, which runs over UDP, and uses TLS to secure the connection. | Authentication credentials, Address, Port, and SNI. Preserve obfuscation and bandwidth settings if they are specified. | Using settings for a different protocol generation, or overlooking UDP restrictions on the current network. |
The same protocol can use different transport combinations. For example, two VLESS configurations with the same address and UUID are not interchangeable if their security layer, Path, or Flow differs. After importing, open the entry and check its detailed fields; do not rely on its name in Home alone.
How to check: verify the original settings before comparing protocols
Use the address, port, and transport settings in your existing service configuration as a baseline. If the connection fails, correct one discrepancy at a time rather than changing the protocol, port, and TLS settings together. This makes it easier to identify the field causing the problem.
Protocol, transport, and TLS settings are different
Type specifies which protocol to use when interpreting the configuration. Transport specifies the channel that carries the data. TLS covers the encrypted handshake and server identity verification. These are separate layers of settings. Changing Type to Trojan does not convert an existing VMess service to Trojan; the remote server must actually provide a Trojan service.
- Address and Port: These specify the connection endpoint. If your existing settings specify port
8443, keep8443; do not replace it with443just because that port appears in common examples. - SNI: Specifies the target server name during the TLS handshake. Address can be an IP while SNI remains the domain name on the certificate. Follow your original settings.
- Host and Path: These may be required with transports such as WebSocket, for example the sample path
/gateway. They are not the same as SNI. Preserve the Path's capitalization and leading slash. - UUID and Password: These are different credential types used by different protocols. Do not substitute one for the other just because both look like strings, and avoid adding spaces or line breaks when copying them.
- Allow Insecure: Relaxes certificate verification and should not be used as a routine fix. If you see a certificate error, first check SNI, your device's date and time, and the certificate's validity.
WireGuard's public-key system differs from TLS certificates, so Trojan's SNI troubleshooting steps do not apply. Hysteria2 carries QUIC over UDP; a reachable TCP port does not prove that the corresponding UDP path is available. Matching protocol settings is only the first step: the network path and remote listener must also be working.
Import and Check Existing Settings in Home
These steps assume you already have your own service settings or subscription. This site does not provide, sell, or recommend servers or subscriptions. A single-share link and a subscription URL serve different purposes: the former usually describes one connection entry, while the latter points to a collection of settings that can be updated. Do not use them interchangeably just because both can be copied as text.
-
Identify the configuration type
First, distinguish a single share link, a subscription URL, and a complete configuration file. Import a subscription URL through Subscribe; do not enter it in the Address field for a regular server.
-
Add a subscription
Go to Home → tap “+” in the top-right corner → Type → Subscribe. Paste your own subscription URL and save. The example URL
https://example.com/sub?token=xxxxillustrates the format only; it is not a working service. -
Check the entry
After updating the subscription, open the entry you need from Home and edit it. Confirm that Type matches the original settings, then check the address, port, credentials, TLS, and transport options one by one.
-
Connect
Select the entry to test in Home, check Global Routing, then turn on the connection. When the system first asks to add a VPN configuration, confirm the app and the source of the request, then follow the system prompts to authorize it.
-
Test and record the results
Run Connectivity Test from Home and open the destination you need to access. Record the current network, selected entry, Global Routing setting, and any errors. Change only one setting at a time before testing again.
To add a server manually, tap “+” in the top-right corner of Home, select Type, then fill in the relevant fields. A subscription update may overwrite entries under that subscription, so do not rely on a temporary manual edit as a permanent fix. If the remote service keeps supplying incorrect settings, check the source of your subscription. Subscription URLs often contain access credentials; hide tokens, passwords, and private keys in screenshots and logs.
Use Global Routing to Separate Rule Issues from Connection Issues
“The protocol was imported” and “the request to a destination uses that protocol” are two different things. If a server is selected in Home but Global Routing is set to Direct, you cannot use website access results to judge whether that server works. The options below describe Shadowrocket's routing modes. Use Config only after checking that the rules file is enabled and suits your needs.
Config
RecommendedSelect a policy based on the rules in the current configuration file. Requests matching DIRECT connect directly; those matching a proxy policy use the corresponding connection. The final policy handles requests that do not match earlier rules.
Best for: everyday traffic routing after verifying your rules
Proxy
Route app traffic through the proxy. This can temporarily reduce the effect of rule matching during troubleshooting, but it will not fix an incorrect password, certificate hostname, or unreachable remote port.
Best for: briefly checking the selected server and destination
Direct
Use a direct connection. Do not treat website accessibility in this mode as a test of the server. Keep in mind that switching modes changes the path used to reach the destination.
Best for: checking your local network with destinations that allow direct access
Scene in Global Routing switches behavior according to configured conditions; it is not another proxy protocol. When troubleshooting, first check which settings the active scene uses. For consistent test conditions, explicitly select Proxy or a Config with known rules to avoid misdiagnosis when scene conditions change.
[Rule]
DOMAIN-SUFFIX,example.com,PROXY
IP-CIDR,192.168.0.0/16,DIRECT,no-resolve
GEOIP,CN,DIRECT
FINAL,PROXY
This illustrates rule syntax; it is not a general configuration recommendation. Rules are usually evaluated in order: DOMAIN-SUFFIX selects a policy by domain suffix, IP-CIDR matches an IP range, GEOIP matches IP geolocation, and FINAL handles requests that do not match earlier rules. In this example, FINAL,PROXY means “send all other requests through the proxy,” so Config cannot be reduced to “unmatched requests go direct.” Edit rules in Config, and confirm that the intended configuration is enabled before testing.
Find the Setting Behind the Failure
Do not conclude that a protocol is unsuitable for your device based on a single timeout. The same symptom can result from credentials, DNS resolution, network restrictions, or routing choices. Keep the original settings, then check the issues below in order. Where possible, keep the network and destination unchanged during testing.
Why can't I connect after importing a subscription?
A successful import only means the content could be parsed. Go to Home → edit the relevant entry and check Type, Address, Port, and authentication fields. Then confirm that Global Routing is not set to Direct. Finally, run Connectivity Test and try accessing the destination, recording each result separately.
Trojan reports a certificate error. Should I enable Allow Insecure?
Keep certificate verification enabled. Check that SNI matches the original settings, then check your device's date and time and the server certificate. If the connection address is an IP, pay particular attention to the certificate hostname. Do not leave verification disabled to hide a hostname mismatch or expired certificate.
Why does Hysteria2 time out when other entries connect?
First check Hysteria2 authentication, the UDP port, and SNI, then try again on another network where testing is allowed. QUIC relies on UDP, so working TCP-based connections do not rule out UDP restrictions. Recording the results on both networks is more useful than repeatedly changing bandwidth settings.
Why is there no traffic even though the WireGuard public key is correct?
Check the local Private Key, peer Public Key, Endpoint, and assigned tunnel address together. Confirm that the server has registered the public key for this device. If only some destinations fail, check routing, DNS, and MTU; do not confuse key issues with network-path issues.
Why does the connection turn on again after I close the app? Is the protocol wrong?
Go to Settings → On Demand and check the connection conditions. Automatic activation is a connection-management behavior; it does not mean the protocol has changed. For troubleshooting, temporarily disable the relevant triggers, test on a consistent network, then restore them and verify the expected behavior.
Choosing a Protocol: What Performance Tests Can Tell You
Protocol availability depends first on what the remote service provides. If your existing settings provide only Trojan, switching locally to WireGuard will not make the same address compatible. Compare performance only when you have valid configurations for multiple protocols, using the same network and destination.
- Keep test conditions consistent: Use the same device, network, and destination. Record whether the connection works, its initial response, and sustained transfer behavior; do not judge by the entry name or a single latency result.
- Distinguish transport capabilities: A protocol that supports UDP does not mean the current network allows UDP. TLS support does not mean certificate settings can be omitted.
- Test the actual use case: A successful Connectivity Test does not guarantee access to every website, streaming service, or UDP app. Verify the services you need separately.
- Keep the original settings: Record the original values before making changes and restore them if performance gets worse. Do not change the protocol, transport, security layer, and rules all at once, or you will not know which change affected the result.
For these reasons, this guide does not rank protocols from fastest to slowest. Device load, packet loss, server resources, destination routes, and specific settings all affect performance. Identifying the protocol helps you find the right fields to check and narrow down problems; the protocol name alone cannot tell you how well an unknown connection will perform.