diff --git a/plugins/gzip.lua b/plugins/gzip.lua
new file mode 100644
index 0000000..8cc93a1
--- /dev/null
+++ b/plugins/gzip.lua
@@ -0,0 +1,185 @@
+--
+-- gzip.lua: GZIP File Format
+--
+-- Copyright (c) 2017, Přemysl Janouch <p.janouch@gmail.com>
+--
+-- Permission to use, copy, modify, and/or distribute this software for any
+-- purpose with or without fee is hereby granted, provided that the above
+-- copyright notice and this permission notice appear in all copies.
+--
+-- THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+-- WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+-- MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY
+-- SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+-- WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION
+-- OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN
+-- CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+--
+
+local detect = function (c)
+	return c:read (2) == "\x1f\x8b"
+end
+
+local function latin1_to_utf8 (s)
+	local u = ""
+	for _, c in ipairs (table.pack (s:byte (1, #s))) do
+		if c < 0x80 then
+			u = u .. string.char (c)
+		else
+			u = u .. string.char (0xc0 | c >> 6, 0x80 | c & 0x3f)
+		end
+	end
+	return u
+end
+
+-- Everything here is based on RFC 1952 and some bits of dictzip
+local crc32_table = function ()
+	local table = {}
+	for n = 0, 255 do
+		local c = n
+		for k = 0, 7 do
+			if c & 1 ~= 0 then
+				c = 0xedb88320 ~ (c >> 1)
+			else
+				c = c >> 1
+			end
+		end
+		table[n] = c
+	end
+	return table
+end
+
+local crc32 = function (s)
+	local table, c = crc32_table (), 0xffffffff
+	for n = 1, #s do c = table[(c ~ s:byte (n)) & 0xff] ~ (c >> 8) end
+	return c ~ 0xffffffff
+end
+
+local os_table = {
+	[0]   = "FAT filesystem",
+	[1]   = "Amiga",
+	[2]   = "VMS",
+	[3]   = "Unix",
+	[4]   = "VM/CMS",
+	[5]   = "Atari TOS",
+	[6]   = "HPFS filesystem",
+	[7]   = "Macintosh",
+	[8]   = "Z-System",
+	[9]   = "CP/M",
+	[10]  = "TOPS-20",
+	[11]  = "NTFS filesystem",
+	[12]  = "QDOS",
+	[13]  = "Acord RISCOS",
+	[255] = "unknown"
+}
+
+local decode = function (c)
+	if not detect (c ()) then error ("not a GZIP file") end
+	local start = c.position
+
+	c.endianity = 'le'
+	c:u16 ("GZIP magic")
+
+	local deflate
+	c:u8 ("compression method: %s", function (u8)
+		if u8 ~= 8 then return "unknown: %d", u8 end
+		deflate = true
+		return "deflate"
+	end)
+
+	local text, hcrc, extra, name, comment
+	c:u8 ("flags: %s", function (u8)
+		text    =  u8       & 1 == 1
+		hcrc    = (u8 >> 1) & 1 == 1
+		extra   = (u8 >> 2) & 1 == 1
+		name    = (u8 >> 3) & 1 == 1
+		comment = (u8 >> 4) & 1 == 1
+
+		local flags = ""
+		if text    then flags = flags .. ", text"       end
+		if hcrc    then flags = flags .. ", header CRC" end
+		if extra   then flags = flags .. ", extra"      end
+		if name    then flags = flags .. ", filename"   end
+		if comment then flags = flags .. ", comment"    end
+
+		if flags == "" then
+			return "none"
+		else
+			return "%s", flags:sub (3)
+		end
+	end)
+
+	c:u32 ("modified time: %s", function (u32)
+		if u32 == 0 then return "none" end
+		return os.date ("!%F %T", u32)
+	end)
+	c:u8 ("extra flags: %s", function (u8)
+		if deflate then
+			if u8 == 2 then return "slowest (%d)", u8 end
+			if u8 == 4 then return "fastest (%d)", u8 end
+		end
+		return "unknown: %d", u8
+	end)
+	c:u8 ("OS: %s", function (u8)
+		os = os_table[u8]
+		if os then return os end
+		return "unknown: %d", u8
+	end)
+
+	local extra_table = {}
+	if extra then
+		local len = c:u16 ("extra field length: %d")
+		c (c.position, c.position + len - 1):mark ("extra field")
+
+		-- This will handle even overflowing subfields
+		while len >= 4 do
+			local p, sid = c.position, c:read (2)
+			c (p, c.position - 1):mark ("subfield ID: %s", sid)
+			local sid_len = c:u16 ("subfield length: %d")
+
+			local subfield = c (c.position, c.position + sid_len - 1)
+			subfield:mark ("subfield data")
+			extra_table[sid] = subfield
+			c.position = c.position + sid_len
+			len = len - 4 - sid_len
+		end
+		c.position = c.position + len
+	end
+
+	if name then
+		c:cstring ("filename: %s", latin1_to_utf8)
+	end
+	if comment then
+		c:cstring ("comment: %s", latin1_to_utf8)
+	end
+	if hcrc then
+		c:u16 ("CRC-16: %s", function (u16)
+			local crc = 0xffff & crc32 (c (start):read (c.position - 1))
+			if crc == u16 then check = "ok" else check = "failed" end
+			return "%#06x (%s)", u16, check
+		end)
+	end
+
+	-- Compressed data follows immediately
+	-- We can jump through it without decompression in dictzip v1 archives
+	local ra = extra_table["RA"]
+	if not ra then return end
+	local ra_ver = ra:u16 ("RA version: %d")
+	if ra_ver ~= 1 then return end
+
+	local ra_chunk = ra:u16 ("chunk length: %d")
+	local ra_count = ra:u16 ("chunk count: %d")
+	for i = 1, ra_count do
+		local len = ra:u16 ("chunk " .. i .. " compressed length: %d")
+		c (c.position, c.position + len - 1):mark ("chunk " .. i)
+		c.position = c.position + len
+	end
+	-- 1 final, 01 static, 0000000 end of block, padding discarded
+	-- This is the kind of block that dictzip finalizes archives with
+	if c:u16 () & 0x03ff == 0x0003 then
+		c:u32 ("CRC-32: %#010x")
+		c:u32 ("input size: %d")
+	end
+end
+
+hex.register { type="gzip", detect=detect, decode=decode }