diff --git a/tensorflow/lite/python/interpreter.py b/tensorflow/lite/python/interpreter.py
index 49792214732..0bd7e8e04ba 100644
--- a/tensorflow/lite/python/interpreter.py
+++ b/tensorflow/lite/python/interpreter.py
@@ -49,8 +49,18 @@ except ImportError:
 
 @_tf_export('lite.Interpreter')
 class Interpreter(object):
-  """Interpreter interface for TF-Lite Models."""
+  """Interpreter interface for TensorFlow Lite Models.
 
+  This makes the TensorFlow Lite interpreter accessible in Python.
+  It is possible to use this interpreter in a multithreaded Python environment,
+  but you must be sure to call functions of a particular instance from only
+  one thread at a time. So if you want to have 4 threads running different
+  inferences simultaneously, create  an interpreter for each one as thread-local
+  data. Similarly, if you are calling invoke() in one thread on a single
+  interpreter but you want to use tensor() on another thread once it is done,
+  you must use a synchronization primitive between the threads to ensure invoke
+  has returned before calling tensor().
+  """
   def __init__(self, model_path=None, model_content=None):
     """Constructor.
 
@@ -282,7 +292,10 @@ class Interpreter(object):
     """Invoke the interpreter.
 
     Be sure to set the input sizes, allocate tensors and fill values before
-    calling this.
+    calling this. Also, note that this function releases the GIL so heavy
+    computation can be done in the background while the Python interpreter
+    continues. No other function on this object should be called while the
+    invoke() call has not finished.
 
     Raises:
       ValueError: When the underlying interpreter fails raise ValueError.
diff --git a/tensorflow/lite/python/interpreter_wrapper/interpreter_wrapper.cc b/tensorflow/lite/python/interpreter_wrapper/interpreter_wrapper.cc
index 6023587d3b1..eb8a1f81130 100644
--- a/tensorflow/lite/python/interpreter_wrapper/interpreter_wrapper.cc
+++ b/tensorflow/lite/python/interpreter_wrapper/interpreter_wrapper.cc
@@ -125,7 +125,16 @@ PyObject* InterpreterWrapper::AllocateTensors() {
 
 PyObject* InterpreterWrapper::Invoke() {
   TFLITE_PY_ENSURE_VALID_INTERPRETER();
-  TFLITE_PY_CHECK(interpreter_->Invoke());
+
+  // Release the GIL so that we can run multiple interpreters in parallel
+  TfLiteStatus status_code = kTfLiteOk;
+  Py_BEGIN_ALLOW_THREADS;  // To return can happen between this and end!
+  status_code = interpreter_->Invoke();
+  Py_END_ALLOW_THREADS;
+
+  TFLITE_PY_CHECK(
+      status_code);  // don't move this into the Py_BEGIN/Py_End block
+
   Py_RETURN_NONE;
 }