diff --git a/tests/models/decoder_only/language/test_models.py b/tests/models/decoder_only/language/test_models.py
index d35d87459cd..e55a682c04d 100644
--- a/tests/models/decoder_only/language/test_models.py
+++ b/tests/models/decoder_only/language/test_models.py
@@ -29,7 +29,7 @@
     "openbmb/MiniCPM3-4B",
     "Qwen/Qwen-7B-Chat",
     "Qwen/Qwen2.5-0.5B-Instruct",
-    "ehristoforu/Falcon3-MoE-2x7B-Insruct",
+    "TitanML/tiny-mixtral",
 ]
 
 
@@ -83,9 +83,8 @@
         pytest.param("stabilityai/stablelm-3b-4e1t"),  # stablelm
         pytest.param("bigcode/starcoder2-3b"),  # starcoder2
         pytest.param(
-            "ehristoforu/Falcon3-MoE-2x7B-Insruct",  # mixtral
-            marks=[pytest.mark.cpu_model,
-                   large_gpu_mark(min_gb=48)],
+            "TitanML/tiny-mixtral",  # mixtral
+            marks=[pytest.mark.cpu_model],
         )
     ])
 @pytest.mark.parametrize("max_tokens", [32])
diff --git a/tests/models/registry.py b/tests/models/registry.py
index 8b330109d9a..75832d83d72 100644
--- a/tests/models/registry.py
+++ b/tests/models/registry.py
@@ -191,7 +191,7 @@ def check_available_online(
                                                 trust_remote_code=True),
     "MistralForCausalLM": _HfExamplesInfo("mistralai/Mistral-7B-Instruct-v0.1"),
     "MixtralForCausalLM": _HfExamplesInfo("mistralai/Mixtral-8x7B-Instruct-v0.1",  # noqa: E501
-                                          {"falcon3": "ehristoforu/Falcon3-MoE-2x7B-Insruct"}),  # noqa: E501
+                                          {"tiny": "TitanML/tiny-mixtral"}),  # noqa: E501
     "QuantMixtralForCausalLM": _HfExamplesInfo("mistral-community/Mixtral-8x22B-v0.1-AWQ"),  # noqa: E501
     "MptForCausalLM": _HfExamplesInfo("mpt", is_available_online=False),
     "MPTForCausalLM": _HfExamplesInfo("mosaicml/mpt-7b"),