diff --git a/examples/count_tokens.py b/examples/count_tokens.py
new file mode 100644
index 000000000..b6ddf79df
--- /dev/null
+++ b/examples/count_tokens.py
@@ -0,0 +1,5 @@
+from levanter.store import JaggedArrayStore
+
+a = JaggedArrayStore.open("gs://marin-us-central2/tokenized/dolma/algebraic-stack-cc00cf/train/input_ids", dtype=int)
+
+a.data_size
\ No newline at end of file