diff --git a/tests/benchmarks/test_update.py b/tests/benchmarks/test_update.py
new file mode 100644
index 000000000..de4ee28a4
--- /dev/null
+++ b/tests/benchmarks/test_update.py
@@ -0,0 +1,76 @@
+import asyncio
+import random
+
+from tests.testmodels import BenchmarkFewFields, BenchmarkManyFields
+
+
+def test_update_few_fields_with_save(benchmark, few_fields_benchmark_dataset):
+    loop = asyncio.get_event_loop()
+
+    @benchmark
+    def bench():
+        async def _bench():
+            instance = random.choice(few_fields_benchmark_dataset)  # nosec
+            instance.level = random.randint(0, 100)  # nosec
+            instance.text = "updated " + str(random.randint(0, 100))  # nosec
+            await instance.save()
+
+        loop.run_until_complete(_bench())
+
+
+def test_update_many_fields_with_save(
+    benchmark, many_fields_benchmark_dataset, gen_many_fields_data
+):
+    loop = asyncio.get_event_loop()
+
+    @benchmark
+    def bench():
+        async def _bench():
+            instance = random.choice(many_fields_benchmark_dataset)  # nosec
+            rand_val = random.randint(0, 100)  # nosec
+            instance.col_float1 = random.uniform(0, 100)  # nosec
+            instance.col_smallint1 = rand_val
+            instance.col_int1 = rand_val
+            instance.col_bigint1 = rand_val
+            instance.col_char1 = "updated " + str(rand_val)
+            instance.col_text1 = "updated " + str(rand_val)
+            await instance.save()
+
+        loop.run_until_complete(_bench())
+
+
+def test_update_few_fields_with_update(benchmark, few_fields_benchmark_dataset):
+    loop = asyncio.get_event_loop()
+
+    @benchmark
+    def bench():
+        async def _bench():
+            instance = random.choice(few_fields_benchmark_dataset)  # nosec
+            await BenchmarkFewFields.filter(id=instance.id).update(
+                level=random.randint(0, 100),  # nosec
+                text="updated " + str(random.randint(0, 100)),  # nosec
+            )
+
+        loop.run_until_complete(_bench())
+
+
+def test_update_many_fields_with_update(
+    benchmark, many_fields_benchmark_dataset, gen_many_fields_data
+):
+    loop = asyncio.get_event_loop()
+
+    @benchmark
+    def bench():
+        async def _bench():
+            instance = random.choice(many_fields_benchmark_dataset)  # nosec
+            rand_val = random.randint(0, 100)  # nosec
+            await BenchmarkManyFields.filter(id=instance.id).update(
+                col_float1=random.uniform(0, 100),  # nosec
+                col_smallint1=rand_val,
+                col_int1=rand_val,
+                col_bigint1=rand_val,
+                col_char1="updated " + str(rand_val),
+                col_text1="updated " + str(rand_val),
+            )
+
+        loop.run_until_complete(_bench())