diff --git a/poetry.lock b/poetry.lock
index a4a6f862..af01ea91 100644
--- a/poetry.lock
+++ b/poetry.lock
@@ -1575,13 +1575,13 @@ proxy = ["PyJWT (>=2.8.0,<3.0.0)", "apscheduler (>=3.10.4,<4.0.0)", "backoff", "
 
 [[package]]
 name = "llama-cpp-python"
-version = "0.3.5"
+version = "0.3.8"
 description = "Python bindings for the llama.cpp library"
 optional = false
 python-versions = ">=3.8"
 groups = ["main", "dev"]
 files = [
-    {file = "llama_cpp_python-0.3.5.tar.gz", hash = "sha256:f5ce47499d53d3973e28ca5bdaf2dfe820163fa3fb67e3050f98e2e9b58d2cf6"},
+    {file = "llama_cpp_python-0.3.8.tar.gz", hash = "sha256:31c91323b555c025a76a30923cead9f5695da103dd68c15cdbb4509b17f0ed77"},
 ]
 
 [package.dependencies]
@@ -4276,4 +4276,4 @@ type = ["pytest-mypy"]
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.12,<3.13"
-content-hash = "11d90797bbc8dee54226f9c44d922333558c96820801cbe2cf677e313ff58fd0"
+content-hash = "95b1b793171816979edcbf768afd9334e6adac42cd792520d3fa6dfed25228e1"
diff --git a/pyproject.toml b/pyproject.toml
index fdfd9b05..a30d8dc6 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -15,7 +15,7 @@ fastapi = "==0.115.11"
 uvicorn = "==0.34.0"
 structlog = "==25.2.0"
 litellm = "==1.63.0"
-llama_cpp_python = "==0.3.5"
+llama_cpp_python = "==0.3.8"
 cryptography = "==44.0.2"
 sqlalchemy = "==2.0.39"
 aiosqlite = "==0.21.0"
@@ -52,7 +52,7 @@ build = "==1.2.2.post1"
 wheel = "==0.45.1"
 litellm = "==1.63.0"
 pytest-asyncio = "==0.25.3"
-llama_cpp_python = "==0.3.5"
+llama_cpp_python = "==0.3.8"
 scikit-learn = "==1.6.1"
 python-dotenv = "==1.0.1"
 requests = "^2.32.3"