Back to Blog

What Is VLESS REALITY? Handshake Authentication and XTLS Vision

Technical Editorial Team · Published February 19, 2026 · Updated November 10, 2026

What is VLESS REALITY?

VLESS REALITY is a proxy connection setup commonly used on restricted networks. It helps users access the internet through a proxy server, particularly where traditional VPNs are readily identified or disrupted.

It combines two technologies:

  • VLESS: Handles access to websites and internet services through a proxy server.
  • REALITY: Establishes a secure connection and uses a specific connection authentication mechanism to reduce the risk of identifying the proxy server.

For example, in regions with internet restrictions, such as China, some VPN connections may be blocked. VLESS REALITY provides another way to connect.

Its main value goes beyond encrypting data: it makes proxy connections harder to identify through common handshake characteristics and active probing.

Why is REALITY needed?

Conventional VPNs and proxy services generally need to establish a connection to a server first.

Network filtering systems can observe protocols, server addresses, and communication characteristics to identify and restrict some VPN or proxy traffic.

Some systems also actively connect to suspicious servers and inspect their responses to determine whether they offer a proxy service.

REALITY is designed specifically to address these identification methods.

It lets authorized users connect to the proxy server while giving unauthorized connections responses from a real HTTPS website, rather than directly exposing proxy-specific characteristics.

This provides proxy access while reducing the risk that the server reveals its role when handling unfamiliar connections.

What are the advantages of REALITY?

1. Fewer identifying features in proxy connections

REALITY uses TLS connections similar to those of ordinary HTTPS websites, with a specific handshake and authentication design.

This helps reduce some features exposed by traditional proxy connections, making it harder to identify the proxy solely from the initial handshake.

2. Lower risk of proxy discovery through active probing

Some filtering systems actively connect to suspicious servers to look for proxy-specific responses.

REALITY can forward unauthorized connections to a preselected real website.

An unfamiliar connection may therefore receive a normal website response rather than a proxy-specific authentication error.

3. No separate website certificate for the proxy server

Traditional TLS proxy deployments usually require a domain and its TLS certificate.

REALITY establishes trust through its own key-based authentication, without requiring a separate public CA certificate for the proxy endpoint.

This reduces deployment and certificate maintenance work while avoiding some identifying clues from proxy-specific certificates.

4. Compatible clients and routing rules

VLESS REALITY can be used through clients that support it.

For example, users can connect directly to domestic websites and use a proxy server for overseas sites.

Routing is provided by the client. REALITY protects and authenticates the connection between the client and proxy server.

How do VLESS, REALITY, and Vision differ?

These names often appear in the same proxy node, but they have different roles.

TechnologyMain role
VLESSAccess websites and internet services through a proxy server
REALITYEstablish secure connections and reduce proxy-identifying features
XTLS VisionOptimize the handling of certain encrypted traffic, reducing extra overhead and some traffic characteristics

They can be understood as three cooperating parts:

VLESS handles proxy access, REALITY handles connection security and authentication, and Vision optimizes data transfer.

Vision is not required to use REALITY, although VLESS REALITY is often deployed with Vision.

How can an everyday user use VLESS REALITY?

You do not need to understand the implementation of handshakes, keys, or certificates.

The process resembles using a regular VPN:

  1. Install a client that supports VLESS REALITY.
  2. Import the subscription or node configuration supplied by your provider.
  3. Select a server and connect.
  4. Use global proxy mode or smart routing as needed.

Proxy cores supporting the relevant technologies include Xray, sing-box, and Mihomo. Apps build graphical interfaces around these cores.

For everyday users, connection stability, server network quality, and client compatibility matter more than manually studying protocol parameters.

See the 5 VLESS Clients for Windows, Mac, iPhone, and Android in 2026 for Windows, macOS, Android, and iOS options.

In brief: VLESS REALITY provides secure proxy connections while reducing some characteristics that network filters can identify.

It is particularly suitable for users dealing with network restrictions. REALITY primarily addresses connection characteristics and authentication; availability also depends on the server address, network path, and client.


Technical deep dive: REALITY handshake authentication and implementation

The following sections explain REALITY through TLS handshakes, key-based authentication, certificate verification, Target forwarding, and XTLS Vision.

1. ClientHello, Session ID, and Short ID

REALITY preserves the basic TLS 1.3 handshake structure, adding specific handling in ClientHello and certificate verification.

The Xray client uses uTLS to construct ClientHello and can select browser-style handshake fingerprints, such as Chrome.

In Xray's implementation, the ClientHello Session ID carries authenticated encrypted data, including:

  • Client version.
  • Current timestamp.
  • Short ID.
  • Reserved fields.

Short ID is one of the matching conditions configured on the server.

On receiving ClientHello, the server uses its authentication key material to decrypt and verify the Session ID, then checks the client version, timestamp, Short ID, and other conditions.

These checks determine whether the connection comes from an authorized REALITY client.

2. Authentication with X25519, HKDF, and AES-GCM

REALITY combines key agreement and symmetric encryption for handshake authentication.

The client first uses ephemeral X25519 key material and the preconfigured REALITY server public key to compute a shared secret.

It then derives an authentication key through HKDF and uses AES-GCM to encrypt and authenticate the Session ID data.

The server holds the corresponding private key, allowing it to compute the same shared secret and derive the authentication key.

When receiving the client's handshake data, the server:

  1. Extracts the client's key agreement material.
  2. Computes the X25519 shared secret.
  3. Derives the authentication key through HKDF.
  4. Decrypts and verifies the Session ID using AES-GCM.
  5. Checks the client version, timestamp, and Short ID.

Time-difference and version-range checks depend on the relevant server settings.

After REALITY authentication succeeds, the connection proceeds to the proxy flow. VLESS then checks the proxy user identity and processes access requests.

Technical sources: Xray-core REALITY client implementation; REALITY server authentication.

3. Temporary trusted certificates and server identity verification

Ordinary HTTPS connections generally verify server identity through public CA certificate chains.

REALITY uses a different trust mechanism: the client already has the REALITY server public key and confirms server identity through a specific certificate verification process.

In a valid handshake, the server can construct a Temporary Trusted Certificate.

The client checks the authentication information in the certificate and uses the key derived by both sides to verify its relationship to the expected REALITY server.

In the corresponding implementation, this involves an Ed25519 public key and HMAC verification.

Successful verification confirms the REALITY server identity and allows the proxy session to continue.

The client handles certificate results differently:

Certificate resultClient behavior
Valid REALITY temporary trusted certificateConfirm the REALITY server and continue the proxy connection
Real certificate that passes ordinary website verificationExpected proxy session was not established; may enter Spider mode
Invalid certificateReject the connection

Spider mode can send HTTP requests over the real website connection and visit further paths found on its pages.

This mainly handles real website connections after authentication fails; it differs from a normal VLESS proxy session.

Technical sources: Official REALITY explanation; Xray-core client certificate verification.

4. How does Target forwarding address active probing?

Active probing means a network filtering system connects to a suspicious server and infers its purpose from its responses.

With traditional proxy services, unauthorized connections may receive a distinctive error or be closed in a recognizable way.

REALITY offers another processing path.

The server is configured with a real HTTPS website as its Target.

When REALITY authentication succeeds, the server accepts the connection and proceeds to proxy processing.

When authentication fails, the server can forward the connection to Target, letting the real website continue the TLS handshake and respond.

There are two different destinations here:

  • Target: Preconfigured by the server administrator to handle unauthenticated connections.
  • The user's destination website: Specified in the authenticated VLESS request, such as YouTube or another internet service.

They are not the same destination.

The REALITY server does not need the Target website's TLS private key or a local installation of its certificate.

An implementation can also establish the Target connection in advance and choose the subsequent data path according to the authentication result.

This reduces opportunities to discover the proxy directly through authentication-failure responses.

5. How does XTLS Vision handle TLS-in-TLS?

Accessing an HTTPS website through an encrypted proxy usually involves two encrypted connections:

  • HTTPS encryption between the browser and destination website.
  • The secure connection between the client and proxy server.

This structure is called TLS-in-TLS.

Even when contents are encrypted, the lengths and transmission sequence of the inner TLS handshake can affect observable outer traffic characteristics.

XTLS Vision optimizes traffic handling to address this.

TLS identification

Vision can inspect TLS records in the proxy data stream and identify ClientHello, ServerHello, and subsequent application data.

Padding

During the initial connection stage, Vision can pad some data to change handshake length distributions and reduce obvious features caused by fixed message lengths.

Direct Copy

When specific conditions are met, Vision can end the relevant Padding stage and use a more direct data copying and forwarding path.

Xray's XtlsFilterTls identifies eligible TLS traffic. When TLS version, cipher suite, record completeness, processing state, and other requirements are met, Vision can issue CommandPaddingDirect to switch data processing paths.

This can reduce some additional wrapping, unwrapping, and memory copying overhead.

Vision does not remove HTTPS encryption between the browser and destination website, nor does it simply skip proxy-layer encryption for all data.

Technical source: Xray-core XTLS Vision data processing.

6. REALITY's identification limits and technical trade-offs

REALITY primarily improves handshake authentication, certificate handling, and responses to unauthenticated connections.

Network observers can still obtain some connection information, including:

  • Server IP address and port.
  • Unencrypted parts of the TLS handshake.
  • Packet lengths and transmission timing distributions.
  • Connection duration and traffic direction.

Its effectiveness therefore varies by identification method.

Passive traffic identification

REALITY's handshake design reduces some proxy connection characteristics. XTLS Vision can further adjust initial length distributions for eligible traffic.

Active probing

Authentication and Target forwarding can direct unauthorized connections to a real website's response path.

Server IP blocking

REALITY cannot automatically restore reachability to a server IP blocked at the network layer.

If the server address cannot establish a TCP connection, network paths and address reachability must be addressed first.

7. How does REALITY compare with other connection technologies?

REALITY, ordinary TLS, and Hysteria 2 have different technical roles.

REALITY and ordinary TLS

Both involve secure TLS connections.

A conventional TLS proxy generally establishes server identity through a public CA certificate.

REALITY verifies server identity using preconfigured keys and specific authentication, and adds Target handling for unauthenticated connections.

See VLESS REALITY vs TLS: What's the Difference? for a detailed comparison.

REALITY and Hysteria 2

Common VLESS + REALITY + RAW setups use TCP, while Hysteria 2 uses QUIC/UDP.

Their differences include the underlying transport as well as encryption and authentication.

Where UDP is restricted, a TCP-based setup may be easier to connect. Under some loss or congestion conditions, QUIC's transport mechanisms may behave differently.

See Hysteria 2 (HY2) vs VLESS REALITY: Which Protocol Is Better for Speed, Stability, and Restricted Networks? for more.

Summary

VLESS REALITY separates proxy requests from connection security.

VLESS handles proxy user identity and access requests. REALITY establishes a secure connection, verifies server identity, and handles unauthorized connections.

XTLS Vision optimizes the wrapping and forwarding of TLS traffic where applicable.

Together, they provide normal proxy access while improving connection characteristics against some protocol identification and active probing methods.

For users who need stable connections on restricted networks, VLESS REALITY is a representative technical approach in the Xray ecosystem.