diff --git a/api/README.md b/api/README.md
index 7ee8e8e3f19..fa29d149706 100644
--- a/api/README.md
+++ b/api/README.md
@@ -45,7 +45,7 @@ For testing the current nightly build, use the following:
ai.djl
api
- 0.27.0-SNAPSHOT
+ 0.28.0-SNAPSHOT
```
diff --git a/docs/get.md b/docs/get.md
index 8c4c34502ad..2c6e8b99968 100644
--- a/docs/get.md
+++ b/docs/get.md
@@ -99,7 +99,7 @@ dependencies {
implementation platform("ai.djl:bom:-SNAPSHOT")
}
```
-Currently, the ` = 0.21.0`.
+Currently, the ` = 0.28.0`.
This snapshot version is the same as the custom DJL repository.
You also need to change directory to `djl/bom`. Then build and publish it to maven local same as what was done in `djl`.
diff --git a/examples/pom.xml b/examples/pom.xml
index cc18358e947..e6a09987174 100644
--- a/examples/pom.xml
+++ b/examples/pom.xml
@@ -5,12 +5,12 @@
ai.djl
examples
- 0.27.0-SNAPSHOT
+ 0.28.0-SNAPSHOT
11
11
- 0.27.0-SNAPSHOT
+ 0.28.0-SNAPSHOT
ai.djl.examples.inference.ObjectDetection
diff --git a/gradle.properties b/gradle.properties
index e3d08320a55..826d9ad1af9 100644
--- a/gradle.properties
+++ b/gradle.properties
@@ -11,7 +11,7 @@ systemProp.org.gradle.internal.http.connectionTimeout=60000
# FIXME: Workaround gradle publish issue: https://github.com/gradle/gradle/issues/11308
systemProp.org.gradle.internal.publish.checksums.insecure=true
-djl_version=0.27.0
+djl_version=0.28.0
mxnet_version=1.9.1
pytorch_version=2.1.1
tensorflow_version=2.10.1