diff --git a/vllm/model_executor/models/interfaces.py b/vllm/model_executor/models/interfaces.py
index 607ff55835f1d..1e5d80dd2f313 100644
--- a/vllm/model_executor/models/interfaces.py
+++ b/vllm/model_executor/models/interfaces.py
@@ -111,13 +111,7 @@ class SupportsMultiModal(Protocol):
             the appearances of their corresponding multimodal data item in the
             input prompt.
         """
-        if hasattr(self, "get_multimodal_embeddings"):
-            logger.warning_once(
-                "`get_multimodal_embeddings` for vLLM models is deprecated and will be "
-                "removed in v0.13.0 or v1.0.0, whichever is earlier. Please rename "
-                "this method to `embed_multimodal`."
-            )
-            return self.get_multimodal_embeddings(**kwargs)
+        ...
 
     def get_language_model(self) -> VllmModel:
         """
@@ -196,12 +190,7 @@ class SupportsMultiModal(Protocol):
         if multimodal_embeddings is None or len(multimodal_embeddings) == 0:
             return inputs_embeds
 
-        if is_multimodal is None:
-            raise ValueError(
-                "`embed_input_ids` now requires `is_multimodal` arg, "
-                "please update your model runner according to "
-                "https://github.com/vllm-project/vllm/pull/16229."
-            )
+        assert is_multimodal is not None
 
         return _merge_multimodal_embeddings(
             inputs_embeds=inputs_embeds,
diff --git a/vllm/model_executor/models/interfaces_base.py b/vllm/model_executor/models/interfaces_base.py
index e8d521ec2e8aa..f988873c9c77c 100644
--- a/vllm/model_executor/models/interfaces_base.py
+++ b/vllm/model_executor/models/interfaces_base.py
@@ -68,15 +68,6 @@ def _check_vllm_model_init(model: type[object] | object) -> bool:
 def _check_vllm_model_embed_input_ids(model: type[object] | object) -> bool:
     model_embed_input_ids = getattr(model, "embed_input_ids", None)
     if not callable(model_embed_input_ids):
-        model_get_input_embeddings = getattr(model, "get_input_embeddings", None)
-        if callable(model_get_input_embeddings):
-            logger.warning(
-                "`get_input_embeddings` for vLLM models is deprecated and will be "
-                "removed in v0.13.0 or v1.0.0, whichever is earlier. Please rename "
-                "this method to `embed_input_ids`."
-            )
-            model.embed_input_ids = model_get_input_embeddings
-            return True
         logger.warning(
             "The model (%s) is missing the `embed_input_ids` method.",
             model,
diff --git a/vllm/model_executor/models/mistral_large_3_eagle.py b/vllm/model_executor/models/mistral_large_3_eagle.py
index e3ca9e4ca82d0..37cd4324e53d9 100644
--- a/vllm/model_executor/models/mistral_large_3_eagle.py
+++ b/vllm/model_executor/models/mistral_large_3_eagle.py
@@ -18,15 +18,10 @@ from vllm.model_executor.models.deepseek_v2 import (
     DeepseekV2DecoderLayer,
     DeepseekV2Model,
 )
-from vllm.model_executor.models.interfaces import MultiModalEmbeddings
 from vllm.model_executor.models.mistral_large_3 import MistralLarge3ForCausalLM
-from vllm.multimodal.inputs import NestedTensors
 
-from .utils import (
-    _merge_multimodal_embeddings,
-    make_empty_intermediate_tensors_factory,
-    maybe_prefix,
-)
+from .interfaces import SupportsMultiModal
+from .utils import make_empty_intermediate_tensors_factory, maybe_prefix
 
 logger = init_logger(__name__)
 
@@ -117,26 +112,10 @@ class EagleMistralLarge3ForCausalLM(MistralLarge3ForCausalLM):
         )
         super().__init__(vllm_config=vllm_config, prefix=prefix)
 
-    def get_input_embeddings(
-        self,
-        input_ids: torch.Tensor,
-        multimodal_embeddings: MultiModalEmbeddings | None = None,
-        *,
-        is_multimodal: torch.Tensor | None = None,
-        handle_oov_mm_token: bool = False,
-    ) -> torch.Tensor:
-        inputs_embeds = super().embed_input_ids(input_ids)
+    def get_language_model(self) -> torch.nn.Module:
+        return self.model
 
-        if multimodal_embeddings is None or len(multimodal_embeddings) == 0:
-            return inputs_embeds
-
-        assert is_multimodal is not None
-
-        return _merge_multimodal_embeddings(
-            inputs_embeds=inputs_embeds,
-            multimodal_embeddings=multimodal_embeddings,
-            is_multimodal=is_multimodal,
-        )
+    embed_input_ids = SupportsMultiModal.embed_input_ids  # type: ignore
 
     def forward(
         self,
@@ -155,11 +134,3 @@ class EagleMistralLarge3ForCausalLM(MistralLarge3ForCausalLM):
             "model.embed_tokens.weight",
             "lm_head.weight",
         }
-
-    def embed_input_ids(
-        self,
-        input_ids: torch.Tensor,
-        multimodal_embeddings: NestedTensors | None = None,
-        is_multimodal: torch.Tensor | None = None,
-    ) -> torch.Tensor:
-        return self.model.embed_input_ids(input_ids)
diff --git a/vllm/model_executor/models/phi3v.py b/vllm/model_executor/models/phi3v.py
index b7ae548069f25..0d39e29dcc97b 100644
--- a/vllm/model_executor/models/phi3v.py
+++ b/vllm/model_executor/models/phi3v.py
@@ -687,12 +687,7 @@ class Phi3VForCausalLM(nn.Module, SupportsMultiModal, SupportsPP, SupportsQuant)
         if multimodal_embeddings is None or len(multimodal_embeddings) == 0:
             return inputs_embeds
 
-        if is_multimodal is None:
-            raise ValueError(
-                "`embed_input_ids` now requires `is_multimodal` arg, "
-                "please update your model runner according to "
-                "https://github.com/vllm-project/vllm/pull/16229."
-            )
+        assert is_multimodal is not None
 
         return _merge_multimodal_embeddings(
             inputs_embeds=inputs_embeds,
diff --git a/vllm/model_executor/models/qwen3_vl.py b/vllm/model_executor/models/qwen3_vl.py
index 1add39d6b0a84..eac3774196a0a 100644
--- a/vllm/model_executor/models/qwen3_vl.py
+++ b/vllm/model_executor/models/qwen3_vl.py
@@ -1572,12 +1572,7 @@ class Qwen3VLForConditionalGeneration(
         if multimodal_embeddings is None or len(multimodal_embeddings) == 0:
             return inputs_embeds
 
-        if is_multimodal is None:
-            raise ValueError(
-                "`embed_input_ids` now requires `is_multimodal` arg, "
-                "please update your model runner according to "
-                "https://github.com/vllm-project/vllm/pull/16229."
-            )
+        assert is_multimodal is not None
 
         if self.use_deepstack:
             (