diff --git a/specification/cognitiveservices/OpenAI.Inference/models/completions.create.cadl b/specification/cognitiveservices/OpenAI.Inference/models/completions.create.cadl
index 8093d171b5b4..fd602657dff8 100644
--- a/specification/cognitiveservices/OpenAI.Inference/models/completions.create.cadl
+++ b/specification/cognitiveservices/OpenAI.Inference/models/completions.create.cadl
@@ -162,10 +162,13 @@ model CompletionsLogProbs {
     text_offset?: int32[];
 }
 
-@doc("Measurment of the amount of tokens used in this request and response")
+@doc("""
+Representation of the token counts processed for a completions request.
+Counts consider all tokens across prompts, choices, choice alternates, best_of generations, and other consumers.
+""")
 model CompletionsUsage {
     @doc("Number of tokens received in the completion")
-    completion_token: int32,
+    completion_tokens: int32,
     @doc("Number of tokens sent in the original request")
     prompt_tokens: int32,
     @doc("Total number of tokens transacted in this request/response")