diff --git a/.github/workflows/rust_ci.yml b/.github/workflows/rust_ci.yml
index dbca2dc..364c2c7 100644
--- a/.github/workflows/rust_ci.yml
+++ b/.github/workflows/rust_ci.yml
@@ -21,9 +21,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: stable
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - uses: Swatinem/rust-cache@v2
 
       - name: Build
diff --git a/.github/workflows/rust_release.yml b/.github/workflows/rust_release.yml
index b883615..41cd400 100644
--- a/.github/workflows/rust_release.yml
+++ b/.github/workflows/rust_release.yml
@@ -12,7 +12,7 @@ jobs:
     steps:
       - name: Checkout
         uses: actions/checkout@v4
-      - uses: dtolnay/rust-toolchain@stable
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - name: Publish to crates.io
         env:
           CRATES_IO_TOKEN: ${{secrets.CRATES_IO_TOKEN}}
diff --git a/.gitignore b/.gitignore
index 6985cf1..dc51894 100644
--- a/.gitignore
+++ b/.gitignore
@@ -12,3 +12,6 @@ Cargo.lock
 
 # MSVC Windows builds of rustc generate these, which store debugging information
 *.pdb
+
+.idea
+.vscode
\ No newline at end of file
diff --git a/.vscode/setting.json b/.vscode/setting.json
deleted file mode 100644
index bbf3889..0000000
--- a/.vscode/setting.json
+++ /dev/null
@@ -1,5 +0,0 @@
-{
-  "rust-analyzer.cargo.features": "all",
-
-  "rust-analyzer.linkedProjects": ["./bindings/python/Cargo.toml"]
-}
diff --git a/Cargo.toml b/Cargo.toml
index 7d4673c..4ebe386 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -25,16 +25,15 @@ categories.workspace = true
 exclude = ["/.github", "/.vscode", "/bindings/**"]
 
 [dependencies]
-tiktoken-rs = { version = "0.5.9", optional = true }
-tokenizers = { version = "0.19.1", features = ["http"], optional = true }
-tree-sitter = "0.22"
-openssl = { version = "0.10", features = ["vendored"] }
+tiktoken-rs = { version = "0.6.0", optional = true }
+tokenizers = { version = "0.21.1", features = ["http"], optional = true }
+tree-sitter = "0.25.3"
 
 [dev-dependencies]
-tree-sitter-go = "0.21"
-tree-sitter-md = "0.2"
-tree-sitter-python = "0.21"
-tree-sitter-rust = "0.21"
+tree-sitter-go = "0.23.4"
+tree-sitter-md = "0.3.2"
+tree-sitter-python = "0.23.6"
+tree-sitter-rust = "0.24.0"
 
 [features]
 tiktoken-rs = ["dep:tiktoken-rs"]
diff --git a/bindings/python/Cargo.toml b/bindings/python/Cargo.toml
index c644128..7bb5906 100644
--- a/bindings/python/Cargo.toml
+++ b/bindings/python/Cargo.toml
@@ -17,11 +17,11 @@ crate-type = ["cdylib"]
 
 [dependencies]
 code-splitter = { path = "../..", features = ["tiktoken-rs", "tokenizers"] }
-pyo3 = { version = "0.22.2", features = ["extension-module"] }
-tiktoken-rs = "0.5.9"
-tokenizers = { version = "0.19.1", features = ["http"] }
-tree-sitter = "0.22"
-tree-sitter-go = "0.21"
-tree-sitter-md = "0.2"
-tree-sitter-python = "0.21"
-tree-sitter-rust = "0.21"
+pyo3 = { version = "0.24.1", features = ["extension-module"] }
+tiktoken-rs = "0.6.0"
+tokenizers = { version = "0.21.1", features = ["http"] }
+tree-sitter = "0.25.3"
+tree-sitter-go = "0.23.4"
+tree-sitter-md = "0.3.2"
+tree-sitter-python = "0.23.6"
+tree-sitter-rust = "0.24.0"
diff --git a/bindings/python/src/language.rs b/bindings/python/src/language.rs
index 327492e..0e6dddd 100644
--- a/bindings/python/src/language.rs
+++ b/bindings/python/src/language.rs
@@ -1,8 +1,11 @@
 use pyo3::prelude::*;
+use std::collections::HashMap;
+use std::hash::Hash;
+use std::sync::OnceLock;
 use tree_sitter::Language as TreeSitterLanguage;
 
 #[pyclass(eq)]
-#[derive(Clone, Copy, Debug, PartialEq)]
+#[derive(Clone, Copy, Debug, Hash, PartialEq)]
 #[non_exhaustive]
 pub enum Language {
     Golang,
@@ -11,13 +14,34 @@ pub enum Language {
     Rust,
 }
 
+impl Eq for Language {}
+
+static LANGUAGES: OnceLock<HashMap<Language, tree_sitter::Language>> = OnceLock::new();
+
 impl Language {
+    fn init_languages() -> HashMap<Language, tree_sitter::Language> {
+        HashMap::from([
+            (
+                Language::Golang,
+                tree_sitter::Language::new(tree_sitter_go::LANGUAGE),
+            ),
+            (
+                Language::Markdown,
+                tree_sitter::Language::new(tree_sitter_md::LANGUAGE),
+            ),
+            (
+                Language::Python,
+                tree_sitter::Language::new(tree_sitter_python::LANGUAGE),
+            ),
+            (
+                Language::Rust,
+                tree_sitter::Language::new(tree_sitter_rust::LANGUAGE),
+            ),
+        ])
+    }
+
     pub fn as_tree_sitter_language(&self) -> TreeSitterLanguage {
-        match self {
-            Language::Golang => tree_sitter_go::language(),
-            Language::Markdown => tree_sitter_md::language(),
-            Language::Python => tree_sitter_python::language(),
-            Language::Rust => tree_sitter_rust::language(),
-        }
+        let languages = LANGUAGES.get_or_init(Self::init_languages);
+        languages.get(self).unwrap().clone()
     }
 }
diff --git a/src/sizer/chars.rs b/src/sizer/chars.rs
index 41d9520..06bdce8 100644
--- a/src/sizer/chars.rs
+++ b/src/sizer/chars.rs
@@ -4,10 +4,10 @@ use crate::sizer::Sizer;
 /// A marker struct for counting characters in code chunks.
 ///
 /// ```
+/// use tree_sitter::Language;
 /// use code_splitter::{CharCounter, Splitter};
 ///
-/// let lang = tree_sitter_md::language();
-/// let splitter = Splitter::new(lang, CharCounter).unwrap();
+/// let splitter = Splitter::new(Language::new(tree_sitter_md::LANGUAGE), CharCounter).unwrap();
 ///
 /// let code = b"hello, world!";
 /// let chunks = splitter.split(code).unwrap();
diff --git a/src/sizer/words.rs b/src/sizer/words.rs
index af18f7a..3cceacd 100644
--- a/src/sizer/words.rs
+++ b/src/sizer/words.rs
@@ -4,10 +4,10 @@ use crate::sizer::Sizer;
 /// A marker struct for counting words in code chunks.
 ///
 /// ```
+/// use tree_sitter::Language;
 /// use code_splitter::{Splitter, WordCounter};
 ///
-/// let lang = tree_sitter_md::language();
-/// let splitter = Splitter::new(lang, WordCounter).unwrap();
+/// let splitter = Splitter::new(Language::new(tree_sitter_md::LANGUAGE), WordCounter).unwrap();
 ///
 /// let code = b"hello, world!";
 /// let chunks = splitter.split(code).unwrap();
diff --git a/src/splitter.rs b/src/splitter.rs
index 8b29504..bb4b7f2 100644
--- a/src/splitter.rs
+++ b/src/splitter.rs
@@ -26,19 +26,19 @@ where
     ///
     /// # Example: split by characters
     /// ```
+    /// use tree_sitter::Language;
     /// use code_splitter::{CharCounter, Splitter};
     ///
-    /// let lang = tree_sitter_md::language();
-    /// let splitter = Splitter::new(lang, CharCounter).unwrap();
+    /// let splitter = Splitter::new(Language::new(tree_sitter_md::LANGUAGE), CharCounter).unwrap();
     /// let chunks = splitter.split(b"hello, world!").unwrap();
     /// ```
     ///
     /// # Example: split by words
     /// ```
+    /// use tree_sitter::Language;
     /// use code_splitter::{Splitter, WordCounter};
     ///
-    /// let lang = tree_sitter_md::language();
-    /// let splitter = Splitter::new(lang, WordCounter).unwrap();
+    /// let splitter = Splitter::new(Language::new(tree_sitter_md::LANGUAGE), WordCounter).unwrap();
     /// let chunks = splitter.split(b"hello, world!").unwrap();
     /// ```
     ///
@@ -46,12 +46,12 @@ where
     /// ```
     /// # #[cfg(feature = "tokenizers")]
     /// # {
+    /// use tree_sitter::Language;
     /// use code_splitter::Splitter;
     /// use tokenizers::Tokenizer;
     ///
-    /// let lang = tree_sitter_md::language();
     /// let tokenizer = Tokenizer::from_pretrained("bert-base-cased", None).unwrap();
-    /// let splitter = Splitter::new(lang, tokenizer).unwrap();
+    /// let splitter = Splitter::new(Language::new(tree_sitter_md::LANGUAGE), tokenizer).unwrap();
     /// let chunks = splitter.split(b"hello, world!").unwrap();
     /// # }
     /// ```
@@ -60,12 +60,12 @@ where
     /// ```
     /// # #[cfg(feature = "tiktoken-rs")]
     /// # {
+    /// use tree_sitter::Language;
     /// use code_splitter::Splitter;
     /// use tiktoken_rs::cl100k_base;
     ///
-    /// let lang = tree_sitter_md::language();
     /// let bpe = cl100k_base().unwrap();
-    /// let splitter = Splitter::new(lang, bpe).unwrap();
+    /// let splitter = Splitter::new(Language::new(tree_sitter_md::LANGUAGE), bpe).unwrap();
     /// let chunks = splitter.split(b"hello, world!").unwrap();
     /// # }
     /// ```
@@ -84,10 +84,10 @@ where
     ///
     /// # Example: set the maximum size to 256
     /// ```
+    /// use tree_sitter::Language;
     /// use code_splitter::{CharCounter, Splitter};
     ///
-    /// let lang = tree_sitter_md::language();
-    /// let splitter = Splitter::new(lang, CharCounter)
+    /// let splitter = Splitter::new(Language::new(tree_sitter_md::LANGUAGE), CharCounter)
     ///   .unwrap()
     ///   .with_max_size(256);
     /// let chunks = splitter.split(b"hello, world!").unwrap();
diff --git a/tests/golang.rs b/tests/golang.rs
index 00a1bd3..69fc459 100644
--- a/tests/golang.rs
+++ b/tests/golang.rs
@@ -1,5 +1,6 @@
 use code_splitter::{CharCounter, Sizer, Splitter, WordCounter};
 use std::fs;
+use tree_sitter::Language;
 
 const TEST_FILE: &str = "tests/testdata/rectangle.go";
 
@@ -12,9 +13,8 @@ where
     T: Sizer,
 {
     let code = read_test_file();
-    let lang = tree_sitter_go::language();
 
-    let splitter = Splitter::new(lang, sizer)
+    let splitter = Splitter::new(Language::new(tree_sitter_go::LANGUAGE), sizer)
         .expect("Failed to create golang splitter")
         .with_max_size(max_size);
     let chunks = splitter.split(&code).expect("Failed to split golang code");
diff --git a/tests/markdown.rs b/tests/markdown.rs
index 2361577..863521e 100644
--- a/tests/markdown.rs
+++ b/tests/markdown.rs
@@ -1,5 +1,6 @@
 use code_splitter::{CharCounter, Sizer, Splitter, WordCounter};
 use std::fs;
+use tree_sitter::Language;
 
 const TEST_FILE: &str = "tests/testdata/markdown.md";
 
@@ -12,9 +13,8 @@ where
     T: Sizer,
 {
     let code = read_test_file();
-    let lang = tree_sitter_md::language();
 
-    let splitter = Splitter::new(lang, sizer)
+    let splitter = Splitter::new(Language::new(tree_sitter_md::LANGUAGE), sizer)
         .expect("Failed to create markdown splitter")
         .with_max_size(max_size);
     let chunks = splitter
diff --git a/tests/python.rs b/tests/python.rs
index 5058e50..02d5044 100644
--- a/tests/python.rs
+++ b/tests/python.rs
@@ -1,5 +1,6 @@
 use code_splitter::{CharCounter, Sizer, Splitter, WordCounter};
 use std::fs;
+use tree_sitter::Language;
 
 const TEST_FILE: &str = "tests/testdata/rectangle.py";
 
@@ -12,9 +13,8 @@ where
     T: Sizer,
 {
     let code = read_test_file();
-    let lang = tree_sitter_python::language();
 
-    let splitter = Splitter::new(lang, sizer)
+    let splitter = Splitter::new(Language::new(tree_sitter_python::LANGUAGE), sizer)
         .expect("Failed to create python splitter")
         .with_max_size(max_size);
     let chunks = splitter.split(&code).expect("Failed to split python code");
diff --git a/tests/rust.rs b/tests/rust.rs
index d893ed9..b5b0082 100644
--- a/tests/rust.rs
+++ b/tests/rust.rs
@@ -1,5 +1,6 @@
 use code_splitter::{CharCounter, Sizer, Splitter, WordCounter};
 use std::fs;
+use tree_sitter::Language;
 
 const TEST_FILE: &str = "tests/testdata/rectangle.rs";
 
@@ -12,9 +13,8 @@ where
     T: Sizer,
 {
     let code = read_test_file();
-    let lang = tree_sitter_rust::language();
 
-    let splitter = Splitter::new(lang, sizer)
+    let splitter = Splitter::new(Language::new(tree_sitter_rust::LANGUAGE), sizer)
         .expect("Failed to create rust splitter")
         .with_max_size(max_size);
     let chunks = splitter.split(&code).expect("Failed to split rust code");