diff --git a/src/diffusers/schedulers/scheduling_flow_match_euler_discrete.py b/src/diffusers/schedulers/scheduling_flow_match_euler_discrete.py
index 0e4f5c6a1f97..e1f40807bd20 100644
--- a/src/diffusers/schedulers/scheduling_flow_match_euler_discrete.py
+++ b/src/diffusers/schedulers/scheduling_flow_match_euler_discrete.py
@@ -350,8 +350,11 @@ def set_timesteps(
         else:
             sigmas = self.shift * sigmas / (1 + (self.shift - 1) * sigmas)
 
-        # 3. If required, stretch the sigmas schedule to terminate at the configured `shift_terminal` value
-        if self.config.shift_terminal:
+        # 3. If required, stretch the sigmas schedule to terminate at the configured `shift_terminal` value.
+        # With a single step, the lone sigma is both the first and last point of the schedule, so
+        # `one_minus_z[-1]` is 0 and the stretch factor is a 0/0 division -> nan. Skip stretching in that
+        # case; there is nothing to stretch a single-point schedule against.
+        if self.config.shift_terminal and len(sigmas) > 1:
             sigmas = self.stretch_shift_to_terminal(sigmas)
 
         # 4. If required, convert sigmas to one of karras, exponential, or beta sigma schedules
diff --git a/src/diffusers/schedulers/scheduling_flow_match_lcm.py b/src/diffusers/schedulers/scheduling_flow_match_lcm.py
index 97d4ebbc8e42..8de9ad2199c6 100644
--- a/src/diffusers/schedulers/scheduling_flow_match_lcm.py
+++ b/src/diffusers/schedulers/scheduling_flow_match_lcm.py
@@ -359,8 +359,10 @@ def set_timesteps(
         else:
             sigmas = self.shift * sigmas / (1 + (self.shift - 1) * sigmas)  # type: ignore
 
-        # 3. If required, stretch the sigmas schedule to terminate at the configured `shift_terminal` value
-        if self.config.shift_terminal:
+        # 3. If required, stretch the sigmas schedule to terminate at the configured `shift_terminal` value.
+        # With a single step, the lone sigma is both the first and last point of the schedule, so
+        # `stretch_shift_to_terminal` divides 0 by 0 and produces nan. Skip stretching in that case.
+        if self.config.shift_terminal and len(sigmas) > 1:
             sigmas = self.stretch_shift_to_terminal(sigmas)  # type: ignore
 
         # 4. If required, convert sigmas to one of karras, exponential, or beta sigma schedules
diff --git a/src/diffusers/schedulers/scheduling_unipc_multistep.py b/src/diffusers/schedulers/scheduling_unipc_multistep.py
index 5c2cbcc13ff1..b309b192be7f 100644
--- a/src/diffusers/schedulers/scheduling_unipc_multistep.py
+++ b/src/diffusers/schedulers/scheduling_unipc_multistep.py
@@ -432,7 +432,9 @@ def set_timesteps(
                 sigmas = self.time_shift(mu, 1.0, sigmas)
             else:
                 sigmas = self.config.flow_shift * sigmas / (1 + (self.config.flow_shift - 1) * sigmas)
-            if self.config.shift_terminal:
+            # With a single step, the lone sigma is both the first and last point of the schedule, so
+            # `stretch_shift_to_terminal` divides 0 by 0 and produces nan. Skip stretching in that case.
+            if self.config.shift_terminal and len(sigmas) > 1:
                 sigmas = self.stretch_shift_to_terminal(sigmas)
             eps = 1e-6
             if np.fabs(sigmas[0] - 1) < eps:
diff --git a/tests/schedulers/test_scheduler_flow_match_euler_discrete.py b/tests/schedulers/test_scheduler_flow_match_euler_discrete.py
new file mode 100644
index 000000000000..6a5719aeb87d
--- /dev/null
+++ b/tests/schedulers/test_scheduler_flow_match_euler_discrete.py
@@ -0,0 +1,59 @@
+# Copyright 2026 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import unittest
+
+import torch
+
+from diffusers import FlowMatchEulerDiscreteScheduler
+
+
+class FlowMatchEulerDiscreteSchedulerTest(unittest.TestCase):
+    scheduler_class = FlowMatchEulerDiscreteScheduler
+
+    def get_default_config(self, **kwargs):
+        config = {
+            "num_train_timesteps": 1000,
+            "shift": 3.0,
+        }
+        config.update(**kwargs)
+        return config
+
+    def test_set_timesteps_multi_step_reaches_shift_terminal(self):
+        # Sanity check that the multi-step schedule still stretches to shift_terminal;
+        # guards against the single-step fix below accidentally disabling stretching generally.
+        scheduler = self.scheduler_class(**self.get_default_config(shift_terminal=0.1))
+        scheduler.set_timesteps(num_inference_steps=10)
+        self.assertFalse(torch.isnan(scheduler.sigmas).any())
+        self.assertAlmostEqual(scheduler.sigmas[-2].item(), 0.1, places=5)
+
+    def test_set_timesteps_single_step_with_shift_terminal_is_finite(self):
+        # With num_inference_steps=1 the lone sigma is both the first and last point of the
+        # schedule, so stretch_shift_to_terminal's `one_minus_z[-1] / scale_factor` degenerates to
+        # a 0/0 division and produced nan timesteps/sigmas. See gh-14411.
+        scheduler = self.scheduler_class(**self.get_default_config(shift_terminal=0.1))
+        scheduler.set_timesteps(num_inference_steps=1)
+        self.assertFalse(torch.isnan(scheduler.sigmas).any())
+        self.assertFalse(torch.isnan(scheduler.timesteps).any())
+
+    def test_step_single_step_with_shift_terminal_runs(self):
+        scheduler = self.scheduler_class(**self.get_default_config(shift_terminal=0.1))
+        scheduler.set_timesteps(num_inference_steps=1)
+
+        sample = torch.randn(1, 4, 4, 4)
+        model_output = torch.randn_like(sample)
+
+        prev_sample = scheduler.step(model_output, scheduler.timesteps[0], sample).prev_sample
+        self.assertEqual(prev_sample.shape, sample.shape)
+        self.assertFalse(torch.isnan(prev_sample).any())
diff --git a/tests/schedulers/test_scheduler_unipc.py b/tests/schedulers/test_scheduler_unipc.py
index ac7e1d3f88b4..5e53467df34f 100644
--- a/tests/schedulers/test_scheduler_unipc.py
+++ b/tests/schedulers/test_scheduler_unipc.py
@@ -403,6 +403,15 @@ def test_exponential_sigmas(self):
     def test_flow_and_karras_sigmas(self):
         self.check_over_configs(use_flow_sigmas=True, use_karras_sigmas=True)
 
+    def test_flow_sigmas_single_step_with_shift_terminal_is_finite(self):
+        # With num_inference_steps=1 the lone sigma is both the first and last point of the
+        # schedule, so stretch_shift_to_terminal's `one_minus_z[-1] / scale_factor` degenerates to
+        # a 0/0 division and produced nan timesteps/sigmas. See gh-14411.
+        scheduler = UniPCMultistepScheduler(use_flow_sigmas=True, flow_shift=3.0, shift_terminal=0.1)
+        scheduler.set_timesteps(num_inference_steps=1)
+        self.assertFalse(torch.isnan(scheduler.sigmas).any())
+        self.assertFalse(torch.isnan(scheduler.timesteps).any())
+
     def test_flow_and_karras_sigmas_values(self):
         num_train_timesteps = 1000
         num_inference_steps = 5