diff --git a/targets.bzl b/targets.bzl
index fdd69cf..4fbf61c 100644
--- a/targets.bzl
+++ b/targets.bzl
@@ -14,7 +14,7 @@ def define_common_targets():
         name = "headers",
         exported_headers = subdir_glob([
             ("include", "pytorch/tokenizers/*.h"),
-        ], exclude = ["pcre2_regex.h", "std_regex.h"]),
+        ]),
         visibility = [
             "@EXECUTORCH_CLIENTS",
             "//pytorch/tokenizers/...",
@@ -23,9 +23,6 @@ def define_common_targets():
         platforms = PLATFORMS,
     )
 
-    # TODO: add target for regex which does lookahed with pcre2
-    # by adding "-DSUPPORT_REGEX_LOOKAHEAD" as a compiler flag
-    # and including pcre2 dependencies.
     runtime.cxx_library(
         name = "regex",
         srcs = [
@@ -43,6 +40,27 @@ def define_common_targets():
         platforms = PLATFORMS,
     )
 
+    runtime.cxx_library(
+        name = "regex_lookahead",
+        srcs = [
+            "src/pcre2_regex.cpp",
+            "src/regex.cpp",
+            "src/re2_regex.cpp",
+            "src/std_regex.cpp",
+        ],
+        exported_deps = [
+            ":headers",
+        ],
+        exported_external_deps = [
+            "pcre2",
+            "re2",
+        ],
+        preprocessor_flags = ["-DSUPPORT_REGEX_LOOKAHEAD=ON"],
+        visibility = ["//pytorch/tokenizers/..."],
+        header_namespace = "",
+        platforms = PLATFORMS,
+    )
+
     runtime.cxx_library(
         name = "bpe_tokenizer_base",
         srcs = [
@@ -101,6 +119,29 @@ def define_common_targets():
         platforms = PLATFORMS,
     )
 
+    runtime.cxx_library(
+        name = "tiktoken_lookahead",
+        srcs = [
+            "src/tiktoken.cpp",
+        ],
+        deps = [
+            ":regex_lookahead",
+        ],
+        exported_deps = [
+            ":bpe_tokenizer_base",
+            ":headers",
+        ],
+        exported_external_deps = [
+            "pcre2",
+            "re2",
+        ],
+        visibility = [
+            "@EXECUTORCH_CLIENTS",
+            "//pytorch/tokenizers/...",
+        ],
+        platforms = PLATFORMS,
+    )
+
     runtime.cxx_library(
         name = "hf_tokenizer",
         srcs = [