diff --git a/src/veritext/metrics/__init__.py b/src/veritext/metrics/__init__.py
index 5f235fe..811d41b 100644
--- a/src/veritext/metrics/__init__.py
+++ b/src/veritext/metrics/__init__.py
@@ -10,6 +10,7 @@ from veritext.metrics.results import (
     ReadabilityResult,
     RougeResult,
     RougeScore,
+    SemanticResult,
 )
 from veritext.metrics.rouge import Rouge
 
@@ -26,4 +27,5 @@ __all__ = [
     "Rouge",
     "RougeResult",
     "RougeScore",
+    "SemanticResult",
 ]
diff --git a/src/veritext/metrics/results.py b/src/veritext/metrics/results.py
index 54f86ef..53fcd21 100644
--- a/src/veritext/metrics/results.py
+++ b/src/veritext/metrics/results.py
@@ -91,3 +91,20 @@ class ReadabilityResult(BaseModel):
     def score(self) -> float:
         """Return Flesch reading ease as the primary score."""
         return self.flesch_reading_ease
+
+
+class SemanticResult(BaseModel):
+    """Result of semantic similarity computation."""
+
+    model_config = ConfigDict(frozen=True)
+
+    similarity: float
+    """Cosine similarity score (0.0 to 1.0)."""
+
+    model: str
+    """Name of the embedding model used."""
+
+    @property
+    def score(self) -> float:
+        """Return the primary score for this result."""
+        return self.similarity