JA4 fingerprint

t13d171200_5b57614c22b0_173e58c88f4d

Go - a non-browser HTTP client.

TCP · TLS 1.3, domain present, 17 cipher suites, 12 extensions, none offered ALPN.

Try:

Anatomy

t13d171200_5b57614c22b0_173e58c88f4d
tTransportThe protocol carrying the handshake - t for TCP, q for QUIC (HTTP/3), d for DTLS.TCP
13TLS versionThe highest TLS version the client offered, from the supported_versions extension - 13 is TLS 1.3, 12 is TLS 1.2.TLS 1.3
dSNIWhether the client sent a Server Name Indication - d if it named a domain, i if it connected to a bare IP.domain present
17CiphersHow many cipher suites the client offered (GREASE values excluded). The exact list is hashed into the first JA4 hash.17 suites offered
12ExtensionsHow many TLS extensions the client sent (GREASE excluded, and SNI and ALPN are not counted). The list is hashed into the second JA4 hash.12 present
00ALPNThe first application protocol the client advertised - h2 is HTTP/2, h1 is HTTP/1.1, h3 is HTTP/3, 00 is none.none offered

5b57614c22b0 is a SHA-256 of the sorted cipher list · 173e58c88f4d covers the extensions and signature algorithms. Hashes aren't reversible - the identity below comes from matching them.

Identity

from reference DB
Go
Non-browser HTTP client

This exact JA4 string is in our reference database, so it's a confident, byte-for-byte match for Go.

Family
go
Confidence
Exact match · full JA4
Matched on
full JA4 hash
Verified
aggregated, unconfirmed

Get hundreds of insights

connection_fingerprint: {
ja4: {
hash: "t13d171200_5b57614c22b0_173e58c88f4d",
family: "go",
},
user_agent_alignment: "?", // needs a live request
h2_fingerprint: "?", // needs a live request
}

A JA4 is one clue. Foil checks hundreds more - the network, the device, and how the session behaves - and tells you when they don't add up. That's usually where the bots are.

See it on a real session - free →

What to do

Challenge

A non-browser HTTP library is expected on APIs and server-to-server traffic - allow it there, but authenticate it (API keys or mTLS) and rate-limit per credential.

On human-facing flows - login, signup, checkout, password reset, coupon redemption - a library JA4 behind a browser User-Agent is one of the cheapest high-confidence automation tells: block or challenge it, and rate-limit by IP/ASN to blunt credential-stuffing and scraping. Caveat: impersonation tools like curl-impersonate and tls-client can forge a browser JA4, so use the library fingerprint to flag traffic, not to clear it.

About this fingerprint

Go is a non-browser HTTP client. Its TLS handshake looks nothing like a real browser's, so a request carrying this fingerprint alongside a Chrome or Safari User-Agent is an immediate contradiction - one of the cheapest automation tells there is.

The breakdown above is computed from the string itself: the readable JA4_aprefix gives the transport, TLS version, SNI flag, cipher and extension counts, and ALPN, while the two hashes are matched against Foil's reference database to name the client. For the full background, see TLS fingerprinting, or try any fingerprint of your own in the JA4 database.

Both component hashes can be looked up on their own: cipher hash 5b57614c22b0 lists every fingerprint built on this cipher stack, and extension hash 173e58c88f4d lists the variants that share this extension set.

Frequently asked

JA4 is a fingerprint of the TLS ClientHello a client sends before any HTTP. It is a three-part string, JA4_a_JA4_b_JA4_c: a readable prefix (transport, TLS version, SNI, cipher and extension counts, ALPN) followed by two truncated SHA-256 hashes over the sorted cipher and extension lists. Unlike JA3, it survives Chrome's extension-order randomization.

The JA4_a prefix decodes by eye: the first character is the transport, the next two are the TLS version, then SNI, a two-digit cipher count, a two-digit extension count, and the first ALPN value. The two hashes are not reversible; you identify them by matching against a database or by decoding the raw (JA4_r) variant.

A JA3 fingerprint is an MD5 hash of the handshake, so the hash itself carries no readable structure. Only the raw JA3 string (version, ciphers, extensions, curves, point formats) can be decoded back into its components.

Yes. Fingerprints are matched against a distilled reference database of common browsers, HTTP libraries, and automation tools. The match tells you what a client is; identifying whether it is lying about that requires cross-checking against the User-Agent, HTTP/2 fingerprint, and JavaScript environment in a live request.