diff --git a/docs/src/main/sphinx/release.md b/docs/src/main/sphinx/release.md
index 27241823311e..58a276a13a6f 100644
--- a/docs/src/main/sphinx/release.md
+++ b/docs/src/main/sphinx/release.md
@@ -6,6 +6,7 @@
 ```{toctree}
 :maxdepth: 1
 
+release/release-444
 release/release-443
 release/release-442
 release/release-441
diff --git a/docs/src/main/sphinx/release/release-444.md b/docs/src/main/sphinx/release/release-444.md
new file mode 100644
index 000000000000..f8376f9c91d1
--- /dev/null
+++ b/docs/src/main/sphinx/release/release-444.md
@@ -0,0 +1,108 @@
+# Release 444 (3 Apr 2024)
+
+## General
+
+* Improve planning time for queries with a large number of joins. ({issue}`21360`)
+* Fix failure for queries containing large numbers of `LIKE` terms in boolean
+  expressions. ({issue}`21235`)
+* Fix potential failure when queries contain filtered aggregations. ({issue}`21272`)
+
+## Docker image
+
+* Update Java runtime to Java 22. ({issue}`21161`)
+
+## BigQuery connector
+
+* Fix failure when reading BigQuery views with [Apache
+  Arrow](https://arrow.apache.org/docs/). ({issue}`21337`)
+
+## ClickHouse connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## Delta Lake connector
+
+* Add support for reading `BYTE_STREAM_SPLIT` encoding in Parquet files. ({issue}`8357`)
+* Add support for [Canned ACLs](https://docs.aws.amazon.com/AmazonS3/latest/userguide/acl-overview.html#canned-acl)
+  with the native S3 file system. ({issue}`21176`)
+* Add support for concurrent, non-conflicting writes when a table is read and
+  written to in the same query. ({issue}`20983`)
+* Add support for reading tables with [v2
+  checkpoints](https://github.com/delta-io/delta/blob/master/PROTOCOL.md#v2-spec).
+  ({issue}`19345`)
+* Add support for reading [shallow cloned
+  tables](https://docs.databricks.com/sql/language-manual/delta-clone.html).
+  ({issue}`17011`)
+* {{breaking}} Remove support for split size configuration with the catalog
+  properties `delta.max-initial-splits` and `delta.max-initial-split-size`, and
+  the catalog session property `max_initial_split_size`. ({issue}`21320`)
+* Fix incorrect results when querying a table that's being modified
+  concurrently. ({issue}`21324`)
+
+## Druid connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## Hive connector
+
+* Add support for reading `BYTE_STREAM_SPLIT` encoding in Parquet files. ({issue}`8357`)
+* Add support for [Canned ACLs](https://docs.aws.amazon.com/AmazonS3/latest/userguide/acl-overview.html#canned-acl)
+  with the native S3 file system. ({issue}`21176`)
+
+## Hudi connector
+
+* Add support for reading `BYTE_STREAM_SPLIT` encoding in Parquet files. ({issue}`8357`)
+* Add support for [Canned ACLs](https://docs.aws.amazon.com/AmazonS3/latest/userguide/acl-overview.html#canned-acl)
+  with the native S3 file system. ({issue}`21176`)
+
+## Iceberg connector
+
+* Add support for the `metadata_log_entries` system table. ({issue}`20410`)
+* Add support for reading `BYTE_STREAM_SPLIT` encoding in Parquet files. ({issue}`8357`)
+* Add support for [Canned ACLs](https://docs.aws.amazon.com/AmazonS3/latest/userguide/acl-overview.html#canned-acl)
+  with the native S3 file system. ({issue}`21176`)
+
+## Ignite connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## MariaDB connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## MySQL connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## Oracle connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## PostgreSQL connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## Redshift connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## SingleStore connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## Snowflake connector
+
+* Add support for table comments. ({issue}`21305`)
+* Improve performance of queries with `ORDER BY ... LIMIT` clause, or `avg`,
+  `count(distinct)`, `stddev`, or `stddev_pop` aggregation functions when the
+  computation can be pushed down to the underlying database. ({issue}`21219`,
+  {issue}`21148`, {issue}`21130`, {issue}`21338`)
+* Improve performance of reading table comments.  ({issue}`21161`)
+
+## SQLServer connector
+
+* Improve performance of reading table comments. ({issue}`21238`)
+
+## SPI
+
+* Change group id and capacity of `GroupedAccumulatorState` to `int` type. ({issue}`21333`)