diff --git a/instrumentation/opentelemetry-instrumentation-genai-langchain/.changelog/684.added b/instrumentation/opentelemetry-instrumentation-genai-langchain/.changelog/684.added
new file mode 100644
index 000000000..6722344bb
--- /dev/null
+++ b/instrumentation/opentelemetry-instrumentation-genai-langchain/.changelog/684.added
@@ -0,0 +1 @@
+Capture top_k - `gen_ai.request.top_k` and choice count - `gen_ai.request.choice.count` on chat
diff --git a/instrumentation/opentelemetry-instrumentation-genai-langchain/src/opentelemetry/instrumentation/genai/langchain/callback_handler.py b/instrumentation/opentelemetry-instrumentation-genai-langchain/src/opentelemetry/instrumentation/genai/langchain/callback_handler.py
index ceb4d387d..cdc4043f9 100644
--- a/instrumentation/opentelemetry-instrumentation-genai-langchain/src/opentelemetry/instrumentation/genai/langchain/callback_handler.py
+++ b/instrumentation/opentelemetry-instrumentation-genai-langchain/src/opentelemetry/instrumentation/genai/langchain/callback_handler.py
@@ -291,6 +291,8 @@ def on_chat_model_start(
         request_model = request_model.removeprefix("models/")
 
         # Initialize variables with default values to avoid "possibly unbound" errors
+        request_choice_count = None
+        top_k = None
         top_p = None
         frequency_penalty = None
         presence_penalty = None
@@ -300,6 +302,8 @@ def on_chat_model_start(
         max_tokens = None
 
         if params is not None:
+            request_choice_count = params.get("n")
+            top_k = params.get("top_k")
             top_p = params.get("top_p")
             frequency_penalty = params.get("frequency_penalty")
             presence_penalty = params.get("presence_penalty")
@@ -343,6 +347,8 @@ def on_chat_model_start(
         llm_invocation.conversation_id = _conversation_id(metadata)
         llm_invocation.input_messages = input_messages
         llm_invocation.top_p = top_p
+        llm_invocation.top_k = top_k
+        llm_invocation.request_choice_count = request_choice_count
         llm_invocation.frequency_penalty = frequency_penalty
         llm_invocation.presence_penalty = presence_penalty
         llm_invocation.stop_sequences = stop_sequences
diff --git a/instrumentation/opentelemetry-instrumentation-genai-langchain/tests/test_callback_handler.py b/instrumentation/opentelemetry-instrumentation-genai-langchain/tests/test_callback_handler.py
index c4c0632f0..4fd79a72a 100644
--- a/instrumentation/opentelemetry-instrumentation-genai-langchain/tests/test_callback_handler.py
+++ b/instrumentation/opentelemetry-instrumentation-genai-langchain/tests/test_callback_handler.py
@@ -2799,6 +2799,42 @@ def test_on_chat_model_start_captures_input_messages_when_content_enabled():
     assert blob.content == _REAL_PNG_BYTES
 
 
+def test_on_chat_model_start_captures_top_k_and_choice_count():
+    run_id = _run_id()
+    handler, _, llm_inv = _make_handler_with_llm_invocation(run_id)
+
+    handler.on_chat_model_start(
+        serialized={},
+        messages=[[HumanMessage(content="Hello")]],
+        run_id=run_id,
+        invocation_params={
+            "params": {
+                "model_name": "gpt-4o",
+                "top_k": 40,
+                "n": 3,
+            }
+        },
+    )
+
+    assert llm_inv.top_k == 40
+    assert llm_inv.request_choice_count == 3
+
+
+def test_on_chat_model_start_defaults_top_k_and_choice_count_to_none():
+    run_id = _run_id()
+    handler, _, llm_inv = _make_handler_with_llm_invocation(run_id)
+
+    handler.on_chat_model_start(
+        serialized={},
+        messages=[[HumanMessage(content="Hello")]],
+        run_id=run_id,
+        invocation_params={"model_name": "gpt-4o"},
+    )
+
+    assert llm_inv.top_k is None
+    assert llm_inv.request_choice_count is None
+
+
 def test_on_chat_model_start_preserves_message_name():
     run_id = _run_id()
     handler, telemetry, llm_inv = _make_handler_with_llm_invocation(run_id)