diff --git a/python/tests/test_python_actors.py b/python/tests/test_python_actors.py
index 3f95a8507..c19ac817d 100644
--- a/python/tests/test_python_actors.py
+++ b/python/tests/test_python_actors.py
@@ -85,8 +85,6 @@ async def test_choose():
     assert_type(result, int)
     assert result2 == result3
 
-    await proc.stop()
-
 
 @pytest.mark.timeout(60)
 async def test_stream():
@@ -96,8 +94,6 @@ async def test_stream():
 
     assert 8 == sum([await x for x in v.value.stream()])
 
-    await proc.stop()
-
 
 class To(Actor):
     @endpoint
@@ -120,8 +116,6 @@ async def test_mesh_passed_to_mesh():
     assert len(all) == 4
     assert all[0] != all[1]
 
-    await proc.stop()
-
 
 @pytest.mark.timeout(60)
 async def test_mesh_passed_to_mesh_on_different_proc_mesh():
@@ -133,9 +127,6 @@ async def test_mesh_passed_to_mesh_on_different_proc_mesh():
     assert len(all) == 4
     assert all[0] != all[1]
 
-    await proc.stop()
-    await proc2.stop()
-
 
 @pytest.mark.timeout(60)
 async def test_actor_slicing():
@@ -152,9 +143,6 @@ async def test_actor_slicing():
 
     assert result[0] == result[1]
 
-    await proc.stop()
-    await proc2.stop()
-
 
 @pytest.mark.timeout(60)
 async def test_aggregate():
@@ -165,8 +153,6 @@ async def test_aggregate():
     r = await acc.accumulate()
     assert r == 4
 
-    await proc.stop()
-
 
 class RunIt(Actor):
     @endpoint
@@ -184,8 +170,6 @@ async def test_rank_size():
     assert 1 == await acc.accumulate(lambda: current_rank()["gpus"])
     assert 4 == await acc.accumulate(lambda: current_size()["gpus"])
 
-    await proc.stop()
-
 
 class SyncActor(Actor):
     @endpoint
@@ -201,30 +185,24 @@ async def test_sync_actor():
     r = await a.sync_endpoint.choose(c)
     assert r == 5
 
-    await proc.stop()
-
 
 @pytest.mark.timeout(60)
-async def test_sync_actor_sync_client() -> None:
+def test_sync_actor_sync_client() -> None:
     proc = local_proc_mesh(gpus=2)
     a = proc.spawn("actor", SyncActor).get()
     c = proc.spawn("counter", Counter, 5).get()
     r = a.sync_endpoint.choose(c).get()
     assert r == 5
 
-    await proc.stop()
-
 
 @pytest.mark.timeout(60)
-async def test_proc_mesh_size() -> None:
+def test_proc_mesh_size() -> None:
     proc = local_proc_mesh(gpus=2)
     assert 2 == proc.size("gpus")
-    # proc.initialized.get()
-    # await proc.stop()
 
 
 @pytest.mark.timeout(60)
-async def test_rank_size_sync() -> None:
+def test_rank_size_sync() -> None:
     proc = local_proc_mesh(gpus=2)
     r = proc.spawn("runit", RunIt).get()
 
@@ -232,11 +210,9 @@ async def test_rank_size_sync() -> None:
     assert 1 == acc.accumulate(lambda: current_rank()["gpus"]).get()
     assert 4 == acc.accumulate(lambda: current_size()["gpus"]).get()
 
-    await proc.stop()
-
 
 @pytest.mark.timeout(60)
-async def test_accumulate_sync() -> None:
+def test_accumulate_sync() -> None:
     proc = local_proc_mesh(gpus=2)
     counter = proc.spawn("counter", Counter, 1).get()
     counter.incr.broadcast()
@@ -244,8 +220,6 @@ async def test_accumulate_sync() -> None:
     r = acc.accumulate().get()
     assert r == 4
 
-    await proc.stop()
-
 
 class CastToCounter(Actor):
     @endpoint
@@ -254,7 +228,7 @@ def doit(self, c: Counter):
 
 
 @pytest.mark.timeout(60)
-async def test_value_mesh() -> None:
+def test_value_mesh() -> None:
     proc = local_proc_mesh(gpus=2)
     counter = proc.spawn("counter", Counter, 0).get()
     counter.slice(hosts=0, gpus=1).incr.broadcast()
@@ -265,8 +239,6 @@ async def test_value_mesh() -> None:
     n = proc.spawn("ctc", CastToCounter).get()
     assert list(x) == n.slice(gpus=0).doit.call_one(counter).get()
 
-    await proc.stop()
-
 
 @pytest.mark.timeout(60)
 def test_rust_binding_modules_correct() -> None:
@@ -333,8 +305,6 @@ async def test_actor_tls() -> None:
     assert 4 == await am.get.call_one()
     assert 4 == await am.get_async.call_one()
 
-    await pm.stop()
-
 
 class TLSActorFullSync(Actor):
     """An actor that manages thread-local state."""
@@ -364,8 +334,6 @@ async def test_actor_tls_full_sync() -> None:
 
     assert 4 == await am.get.call_one()
 
-    await pm.stop()
-
 
 class AsyncActor(Actor):
     def __init__(self):
@@ -392,8 +360,6 @@ async def test_async_concurrency():
     await am.no_more.call()
     await fut
 
-    await pm.stop()
-
 
 async def awaitit(f):
     return await f
@@ -750,7 +716,7 @@ async def send(self, port: Port[int]):
 
 
 @pytest.mark.timeout(60)
-async def test_port_as_argument() -> None:
+def test_port_as_argument() -> None:
     proc_mesh = local_proc_mesh(gpus=1)
     s = proc_mesh.spawn("send_alot", SendAlot).get()
     send, recv = Channel[int].open()
@@ -760,8 +726,6 @@ async def test_port_as_argument() -> None:
     for i in range(100):
         assert i == recv.recv().get()
 
-    await proc_mesh.stop()
-
 
 @pytest.mark.timeout(15)
 async def test_same_actor_twice() -> None:
@@ -778,8 +742,6 @@ async def test_same_actor_twice() -> None:
         "gspawn failed: an actor with name 'dup' has already been spawned" in error_msg
     ), f"Expected error message about duplicate actor name, got: {error_msg}"
 
-    await pm.stop()
-
 
 class TestActorMeshStop(unittest.IsolatedAsyncioTestCase):
     async def test_actor_mesh_stop(self) -> None:
@@ -806,13 +768,11 @@ def add(self, port: "Port[int]", b: int) -> None:
 
 
 @pytest.mark.timeout(60)
-async def test_ported_actor():
+def test_ported_actor():
     proc_mesh = local_proc_mesh(gpus=1).get()
     a = proc_mesh.spawn("port_actor", PortedActor).get()
     assert 5 == a.add.call_one(2).get()
 
-    await proc_mesh.stop()
-
 
 async def _recv():
     return (7, 2, 3)