diff --git a/django/contrib/auth/hashers.py b/django/contrib/auth/hashers.py
index 344176be69..7a9084b746 100644
--- a/django/contrib/auth/hashers.py
+++ b/django/contrib/auth/hashers.py
@@ -302,8 +302,8 @@ class Argon2PasswordHasher(BasePasswordHasher):
     library = 'argon2'
 
     time_cost = 2
-    memory_cost = 512
-    parallelism = 2
+    memory_cost = 102400
+    parallelism = 8
 
     def encode(self, password, salt):
         argon2 = self._load_library()
@@ -363,7 +363,7 @@ class Argon2PasswordHasher(BasePasswordHasher):
         argon2 = self._load_library()
         # salt_len is a noop, because we provide our own salt.
         return argon2.Parameters(
-            type=argon2.low_level.Type.I,
+            type=argon2.low_level.Type.ID,
             version=argon2.low_level.ARGON2_VERSION,
             salt_len=argon2.DEFAULT_RANDOM_SALT_LENGTH,
             hash_len=argon2.DEFAULT_HASH_LENGTH,
diff --git a/docs/releases/3.2.txt b/docs/releases/3.2.txt
index 11da13daba..b6db027448 100644
--- a/docs/releases/3.2.txt
+++ b/docs/releases/3.2.txt
@@ -50,6 +50,15 @@ Minor features
 * The default iteration count for the PBKDF2 password hasher is increased from
   216,000 to 260,000.
 
+* The default variant for the Argon2 password hasher is changed to Argon2id.
+  ``memory_cost`` and ``parallelism`` are increased to 102,400 and 8
+  respectively to match the ``argon2-cffi`` defaults.
+
+  Increasing the ``memory_cost`` pushes the required memory from 512 KB to 100
+  MB. This is still rather conservative but can lead to problems in memory
+  constrained environments. If this is the case, the existing hasher can be
+  subclassed to override the defaults.
+
 :mod:`django.contrib.contenttypes`
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
diff --git a/tests/auth_tests/test_hashers.py b/tests/auth_tests/test_hashers.py
index 5ca422764b..fa7332970a 100644
--- a/tests/auth_tests/test_hashers.py
+++ b/tests/auth_tests/test_hashers.py
@@ -497,13 +497,13 @@ class TestUtilsHashPassArgon2(SimpleTestCase):
     def test_argon2(self):
         encoded = make_password('lètmein', hasher='argon2')
         self.assertTrue(is_password_usable(encoded))
-        self.assertTrue(encoded.startswith('argon2$'))
+        self.assertTrue(encoded.startswith('argon2$argon2id$'))
         self.assertTrue(check_password('lètmein', encoded))
         self.assertFalse(check_password('lètmeinz', encoded))
         self.assertEqual(identify_hasher(encoded).algorithm, 'argon2')
         # Blank passwords
         blank_encoded = make_password('', hasher='argon2')
-        self.assertTrue(blank_encoded.startswith('argon2$'))
+        self.assertTrue(blank_encoded.startswith('argon2$argon2id$'))
         self.assertTrue(is_password_usable(blank_encoded))
         self.assertTrue(check_password('', blank_encoded))
         self.assertFalse(check_password(' ', blank_encoded))
@@ -523,15 +523,15 @@ class TestUtilsHashPassArgon2(SimpleTestCase):
 
     def test_argon2_upgrade(self):
         self._test_argon2_upgrade('time_cost', 'time cost', 1)
-        self._test_argon2_upgrade('memory_cost', 'memory cost', 16)
+        self._test_argon2_upgrade('memory_cost', 'memory cost', 64)
         self._test_argon2_upgrade('parallelism', 'parallelism', 1)
 
     def test_argon2_version_upgrade(self):
         hasher = get_hasher('argon2')
         state = {'upgraded': False}
         encoded = (
-            'argon2$argon2i$m=8,t=1,p=1$c29tZXNhbHQ$gwQOXSNhxiOxPOA0+PY10P9QFO'
-            '4NAYysnqRt1GSQLE55m+2GYDt9FEjPMHhP2Cuf0nOEXXMocVrsJAtNSsKyfg'
+            'argon2$argon2id$v=19$m=102400,t=2,p=8$Y041dExhNkljRUUy$TMa6A8fPJh'
+            'CAUXRhJXCXdw'
         )
 
         def setter(password):