diff --git a/Cargo.toml b/Cargo.toml
index de0bd6d..8716ae2 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -12,6 +12,8 @@ homepage = "https://github.com/swaits/bpe-tokenizer/"
 repository = "https://github.com/swaits/bpe-tokenizer/"
 rust-version = "1.64"
 license-file = "LICENSE"
+exclude = ["vocab/multi.wiki.bpe.vs1000000.vocab"]
+# ^ default-large vocab is temporarily excluded until we can get a crates.io exception
 
 [dependencies]
 thiserror = "1.0.64"
diff --git a/README.md b/README.md
index 3c20c29..aa92c41 100644
--- a/README.md
+++ b/README.md
@@ -9,6 +9,9 @@ frequent token pairs.
 It supports **Unicode-aware** text segmentation for sentence and word splitting,
 making it suitable for processing a variety of languages and scripts.
 
+**NOTE:** The `default-large` feature is temporarily disabled due to crate size
+limitations; presently awaiting an exception from <help@crate.io>.
+
 ## Features
 
 - **Bring your own BPE token vocabularies**, or use ...