diff --git a/gallery_dl/extractor/mangafire.py b/gallery_dl/extractor/mangafire.py
index 646729f5..f5285444 100644
--- a/gallery_dl/extractor/mangafire.py
+++ b/gallery_dl/extractor/mangafire.py
@@ -11,7 +11,6 @@
 from .common import ChapterExtractor, MangaExtractor
 from .. import text, exception
 from ..cache import memcache
-import binascii
 
 BASE_PATTERN = r"(?:https?://)?(?:www\.)?mangafire\.to"
 
@@ -53,7 +52,8 @@ class MangafireChapterExtractor(MangafireBase, ChapterExtractor):
 
     def images(self, page):
         url = f"{self.root}/ajax/read/{self.type}/{self.chapter_id}"
-        params = {"vrf": generate_VRF(f"{self.type}@{self.chapter_id}")}
+        params = {"vrf": self.utils("vrf").generate(
+            f"{self.type}@{self.chapter_id}")}
         headers = {"x-requested-with": "XMLHttpRequest"}
         data = self.request_json(url, params=params, headers=headers)
 
@@ -131,7 +131,7 @@ def _manga_info(self, manga_path, page=None):
 def _manga_chapters(self, manga_info):
     manga_id, type, lang = manga_info
     url = f"{self.root}/ajax/read/{manga_id}/{type}/{lang}"
-    params = {"vrf": generate_VRF(f"{manga_id}@{type}@{lang}")}
+    params = {"vrf": self.utils("vrf").generate(f"{manga_id}@{type}@{lang}")}
     headers = {"x-requested-with": "XMLHttpRequest"}
     data = self.request_json(url, params=params, headers=headers)
 
@@ -169,122 +169,3 @@ def _chapter_info(info):
         "title"         : text.unescape(text.extr(info, 'title="', '"')),
         "lang"          : lang,
     }
-
-
-###############################################################################
-# VRF generation utils
-#
-# adapted from dazedcat19/FMD2
-# https://github.com/dazedcat19/FMD2/blob/master/lua/modules/MangaFire.lua
-
-def generate_VRF(input):
-    input = text.quote(input).encode()
-
-    for key_b64, seed_b64, prefix_b64, schedule in (
-        (key_l, seed_A, prefix_O, schedule_c),
-        (key_g, seed_V, prefix_v, schedule_y),
-        (key_B, seed_N, prefix_L, schedule_b),
-        (key_m, seed_P, prefix_p, schedule_j),
-        (key_F, seed_k, prefix_W, schedule_e),
-    ):
-        input = transform(
-            rc4(binascii.a2b_base64(key_b64), input),
-            binascii.a2b_base64(seed_b64),
-            binascii.a2b_base64(prefix_b64),
-            schedule,
-        )
-
-    return binascii.b2a_base64(bytes(input), newline=False).rstrip(
-        b"=").replace(b"+", b"-").replace(b"/", b"_")
-
-
-def transform(input, seed, prefix, schedule):
-    prefix_len = len(prefix)
-
-    out = []
-    for idx, c in enumerate(input):
-        if idx < prefix_len:
-            out.append(prefix[idx] or 0)
-        out.append(schedule[idx % 10]((c ^ seed[idx % 32]) & 255) & 255)
-    return out
-
-
-def rc4(key, input):
-    lkey = len(key)
-
-    j = 0
-    s = list(range(256))
-    for i in range(256):
-        j = (j + s[i] + key[i % lkey]) & 255
-        s[i], s[j] = s[j], s[i]
-
-    out = []
-    i = j = 0
-    for c in input:
-        i = (i + 1) & 255
-        j = (j + s[i]) & 255
-        s[i], s[j] = s[j], s[i]
-        k = s[(s[i] + s[j]) & 255]
-        out.append(c ^ k)
-    return out
-
-
-def add8(n):
-    return lambda c: (c + n) & 255
-
-
-def sub8(n):
-    return lambda c: (c - n + 256) & 255
-
-
-def xor8(n):
-    return lambda c: (c ^ n) & 255
-
-
-def rotl8(n):
-    return lambda c: ((c << n) | (c >> (8 - n))) & 255
-
-
-def rotr8(n):
-    return lambda c: ((c >> n) | (c << (8 - n))) & 255
-
-
-schedule_c = (
-    sub8(223), rotr8(4), rotr8(4), add8(234), rotr8(7),
-    rotr8(2), rotr8(7), sub8(223), rotr8(7), rotr8(6),
-)
-schedule_y = (
-    add8(19), rotr8(7), add8(19), rotr8(6), add8(19),
-    rotr8(1), add8(19), rotr8(6), rotr8(7), rotr8(4),
-)
-schedule_b = (
-    sub8(223), rotr8(1), add8(19), sub8(223), rotl8(2),
-    sub8(223), add8(19), rotl8(1), rotl8(2), rotl8(1),
-)
-schedule_j = (
-    add8(19), rotl8(1), rotl8(1), rotr8(1), add8(234),
-    rotl8(1), sub8(223), rotl8(6), rotl8(4), rotl8(1),
-)
-schedule_e = (
-    rotr8(1), rotl8(1), rotl8(6), rotr8(1), rotl8(2),
-    rotr8(4), rotl8(1), rotl8(1), sub8(223), rotl8(2),
-)
-
-
-key_l = "FgxyJUQDPUGSzwbAq/ToWn4/e8jYzvabE+dLMb1XU1o="
-key_g = "CQx3CLwswJAnM1VxOqX+y+f3eUns03ulxv8Z+0gUyik="
-key_B = "fAS+otFLkKsKAJzu3yU+rGOlbbFVq+u+LaS6+s1eCJs="
-key_m = "Oy45fQVK9kq9019+VysXVlz1F9S1YwYKgXyzGlZrijo="
-key_F = "aoDIdXezm2l3HrcnQdkPJTDT8+W6mcl2/02ewBHfPzg="
-
-seed_A = "yH6MXnMEcDVWO/9a6P9W92BAh1eRLVFxFlWTHUqQ474="
-seed_V = "RK7y4dZ0azs9Uqz+bbFB46Bx2K9EHg74ndxknY9uknA="
-seed_N = "rqr9HeTQOg8TlFiIGZpJaxcvAaKHwMwrkqojJCpcvoc="
-seed_P = "/4GPpmZXYpn5RpkP7FC/dt8SXz7W30nUZTe8wb+3xmU="
-seed_k = "wsSGSBXKWA9q1oDJpjtJddVxH+evCfL5SO9HZnUDFU8="
-
-prefix_O = "l9PavRg="
-prefix_v = "Ml2v7ag1Jg=="
-prefix_L = "i/Va0UxrbMo="
-prefix_p = "WFjKAHGEkQM="
-prefix_W = "5Rr27rWd"
diff --git a/gallery_dl/extractor/utils/mangafire_vrf.py b/gallery_dl/extractor/utils/mangafire_vrf.py
new file mode 100644
index 00000000..92f2d50a
--- /dev/null
+++ b/gallery_dl/extractor/utils/mangafire_vrf.py
@@ -0,0 +1,129 @@
+# -*- coding: utf-8 -*-
+
+# Copyright 2026 Mike Fährmann
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License version 2 as
+# published by the Free Software Foundation.
+
+"""VRF generation utils
+
+adapted from dazedcat19/FMD2
+https://github.com/dazedcat19/FMD2/blob/master/lua/modules/MangaFire.lua
+"""
+
+from ... import text
+import binascii
+
+
+def generate(input):
+    input = text.quote(input).encode()
+
+    for key_b64, seed_b64, prefix_b64, schedule in (
+        (key_l, seed_A, prefix_O, schedule_c),
+        (key_g, seed_V, prefix_v, schedule_y),
+        (key_B, seed_N, prefix_L, schedule_b),
+        (key_m, seed_P, prefix_p, schedule_j),
+        (key_F, seed_k, prefix_W, schedule_e),
+    ):
+        input = transform(
+            rc4(binascii.a2b_base64(key_b64), input),
+            binascii.a2b_base64(seed_b64),
+            binascii.a2b_base64(prefix_b64),
+            schedule,
+        )
+
+    return binascii.b2a_base64(bytes(input), newline=False).rstrip(
+        b"=").replace(b"+", b"-").replace(b"/", b"_")
+
+
+def transform(input, seed, prefix, schedule):
+    prefix_len = len(prefix)
+
+    out = []
+    for idx, c in enumerate(input):
+        if idx < prefix_len:
+            out.append(prefix[idx] or 0)
+        out.append(schedule[idx % 10]((c ^ seed[idx % 32]) & 255) & 255)
+    return out
+
+
+def rc4(key, input):
+    lkey = len(key)
+
+    j = 0
+    s = list(range(256))
+    for i in range(256):
+        j = (j + s[i] + key[i % lkey]) & 255
+        s[i], s[j] = s[j], s[i]
+
+    out = []
+    i = j = 0
+    for c in input:
+        i = (i + 1) & 255
+        j = (j + s[i]) & 255
+        s[i], s[j] = s[j], s[i]
+        k = s[(s[i] + s[j]) & 255]
+        out.append(c ^ k)
+    return out
+
+
+def add8(n):
+    return lambda c: (c + n) & 255
+
+
+def sub8(n):
+    return lambda c: (c - n + 256) & 255
+
+
+def xor8(n):
+    return lambda c: (c ^ n) & 255
+
+
+def rotl8(n):
+    return lambda c: ((c << n) | (c >> (8 - n))) & 255
+
+
+def rotr8(n):
+    return lambda c: ((c >> n) | (c << (8 - n))) & 255
+
+
+schedule_c = (
+    sub8(223), rotr8(4), rotr8(4), add8(234), rotr8(7),
+    rotr8(2), rotr8(7), sub8(223), rotr8(7), rotr8(6),
+)
+schedule_y = (
+    add8(19), rotr8(7), add8(19), rotr8(6), add8(19),
+    rotr8(1), add8(19), rotr8(6), rotr8(7), rotr8(4),
+)
+schedule_b = (
+    sub8(223), rotr8(1), add8(19), sub8(223), rotl8(2),
+    sub8(223), add8(19), rotl8(1), rotl8(2), rotl8(1),
+)
+schedule_j = (
+    add8(19), rotl8(1), rotl8(1), rotr8(1), add8(234),
+    rotl8(1), sub8(223), rotl8(6), rotl8(4), rotl8(1),
+)
+schedule_e = (
+    rotr8(1), rotl8(1), rotl8(6), rotr8(1), rotl8(2),
+    rotr8(4), rotl8(1), rotl8(1), sub8(223), rotl8(2),
+)
+
+
+key_l = "FgxyJUQDPUGSzwbAq/ToWn4/e8jYzvabE+dLMb1XU1o="
+key_g = "CQx3CLwswJAnM1VxOqX+y+f3eUns03ulxv8Z+0gUyik="
+key_B = "fAS+otFLkKsKAJzu3yU+rGOlbbFVq+u+LaS6+s1eCJs="
+key_m = "Oy45fQVK9kq9019+VysXVlz1F9S1YwYKgXyzGlZrijo="
+key_F = "aoDIdXezm2l3HrcnQdkPJTDT8+W6mcl2/02ewBHfPzg="
+
+seed_A = "yH6MXnMEcDVWO/9a6P9W92BAh1eRLVFxFlWTHUqQ474="
+seed_V = "RK7y4dZ0azs9Uqz+bbFB46Bx2K9EHg74ndxknY9uknA="
+seed_N = "rqr9HeTQOg8TlFiIGZpJaxcvAaKHwMwrkqojJCpcvoc="
+seed_P = "/4GPpmZXYpn5RpkP7FC/dt8SXz7W30nUZTe8wb+3xmU="
+seed_k = "wsSGSBXKWA9q1oDJpjtJddVxH+evCfL5SO9HZnUDFU8="
+
+prefix_O = "l9PavRg="
+prefix_v = "Ml2v7ag1Jg=="
+prefix_L = "i/Va0UxrbMo="
+prefix_p = "WFjKAHGEkQM="
+prefix_W = "5Rr27rWd"