diff --git a/.github/workflows/build-docker.yml b/.github/workflows/build-docker.yml
index 5e1644432..5ee52fde9 100644
--- a/.github/workflows/build-docker.yml
+++ b/.github/workflows/build-docker.yml
@@ -11,8 +11,8 @@ on:
 env:
   IMAGE_NAME: ${{ secrets.DOCKERHUB_USERNAME }}/voicevox_engine
   PYTHON_VERSION: '3.8.10'
-  VOICEVOX_CORE_VERSION: '0.10.preview.0'
-  VOICEVOX_CORE_SOURCE_VERSION: '0.10.preview.0'
+  VOICEVOX_CORE_VERSION: '0.10.preview.3'
+  VOICEVOX_CORE_SOURCE_VERSION: '0.10.preview.3'
 
 jobs:
   build-docker:
diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 760711a28..866779aef 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -12,8 +12,8 @@ env:
   IMAGE_NAME: ${{ secrets.DOCKERHUB_USERNAME }}/voicevox_engine
   PYTHON_VERSION: '3.8.10'
   VOICEVOX_RESOURCE_VERSION: '0.10.preview.2'
-  VOICEVOX_CORE_VERSION: '0.10.preview.0'
-  VOICEVOX_CORE_SOURCE_VERSION: '0.10.preview.0'
+  VOICEVOX_CORE_VERSION: '0.10.preview.3'
+  VOICEVOX_CORE_SOURCE_VERSION: '0.10.preview.3'
 
 jobs:
   # Build Mac binary (x64 arch only)
diff --git a/Dockerfile b/Dockerfile
index 885b2401a..a950ea737 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -21,7 +21,7 @@ RUN <<EOF
 EOF
 
 # assert VOICEVOX_CORE_VERSION >= 0.10.preview.0 (ONNX)
-ARG VOICEVOX_CORE_VERSION=0.10.preview.0
+ARG VOICEVOX_CORE_VERSION=0.10.preview.3
 ARG VOICEVOX_CORE_LIBRARY_NAME=libcore_cpu_x64.so
 RUN <<EOF
     set -eux
@@ -203,7 +203,7 @@ COPY --from=download-core-env /opt/voicevox_core /opt/voicevox_core
 COPY --from=download-onnxruntime-env /opt/onnxruntime /opt/onnxruntime
 
 # Install VOICEVOX Core Python module
-ARG VOICEVOX_CORE_SOURCE_VERSION=0.10.preview.0
+ARG VOICEVOX_CORE_SOURCE_VERSION=0.10.preview.3
 RUN <<EOF
     set -eux
 
diff --git a/README.md b/README.md
index 35d9a584b..3243f8706 100644
--- a/README.md
+++ b/README.md
@@ -253,7 +253,7 @@ VOICEVOX ENGINEが使う処理能力を調節したい場合は、CPUスレッ
 
 - 環境変数で指定する
     ```bash
-    CPU_NUM_THREADS=4
+    export VV_CPU_NUM_THREADS=4
     python run.py --voicevox_dir=$VOICEVOX_DIR
     ```
 
diff --git a/run.py b/run.py
index 20fb8d7f2..06341dabb 100644
--- a/run.py
+++ b/run.py
@@ -495,15 +495,17 @@ def speaker_info(speaker_uuid: str):
     parser.add_argument("--voicelib_dir", type=Path, default=None)
     parser.add_argument("--enable_cancellable_synthesis", action="store_true")
     parser.add_argument("--init_processes", type=int, default=2)
-    parser.add_argument("--cpu_num_threads", type=int, default=None)
-    args = parser.parse_args()
-
-    cpu_num_threads: Optional[int] = args.cpu_num_threads
 
     # 引数へcpu_num_threadsの指定がなければ、環境変数をロールします。
     # 環境変数にもない場合は、Noneのままとします。
-    if cpu_num_threads is None:
-        cpu_num_threads = os.getenv("CPU_NUM_THREADS", None)
+    # VV_CPU_NUM_THREADSが空文字列でなく数値でもない場合、エラー終了します。
+    parser.add_argument(
+        "--cpu_num_threads", type=int, default=os.getenv("VV_CPU_NUM_THREADS") or None
+    )
+
+    args = parser.parse_args()
+
+    cpu_num_threads: Optional[int] = args.cpu_num_threads
 
     # voicelib_dir が Noneのとき、音声ライブラリの Python モジュールと同じディレクトリにあるとする
     voicelib_dir: Optional[Path] = args.voicelib_dir