diff --git a/serverless/aws/README.md b/serverless/aws/README.md
index d05c8de..39067c8 100644
--- a/serverless/aws/README.md
+++ b/serverless/aws/README.md
@@ -2,6 +2,8 @@
 
 Generates a Lambda function for deploying PMTiles on Lambda behind Lambda Function URLs or API Gateway. (CloudFront is recommended as a cache in front.)
 
+The `Content-Type` header of responses will be `application/vnd.mapbox-vector-tile` : this is one of the [CloudFront compressible types](https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/ServingCompressedFiles.html#compressed-content-cloudfront-file-types).
+
 ## How To Use
 
 Self-contained Lambda ZIP:
@@ -25,8 +27,6 @@ Configure these Lambda environment variables:
   * Example path setting for objects in a directory: `my_folder/{name}/file.pmtiles`
 * `TILE_PATH`: optional, define the URL route of the tiles API. Default `/{name}/{z}/{x}/{y}.pbf`
 
-For API Gateway integration, your Lambda Proxy Integration route will need to specify a greedy capturing parameter called `proxy` e.g. `/{proxy+}` (the default). API Gateway responses will always be GZIP-encoded.
-
 ## Test Event
 
 Lambda Function URLs:
@@ -54,3 +54,8 @@ TBD
 ### Lambda@Edge
 
 Lambda@Edge's multi-region features have little benefit when fetching data from S3 in a single region, and Lambda@Edge [doesn't support](https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/edge-functions-restrictions.html) environment variables or responses over 1 MB. For globally distributed caching, use CloudFront in combination with Lambda Function URLs.
+
+### API Gateway
+
+* your Lambda Proxy Integration route will need to specify a greedy capturing parameter called `proxy` e.g. `/{proxy+}` (the default).
+* API Gateway responses will always be GZIP-encoded, to work around binary content detection problems.