From 07e677e002ddf645f5a86247040e179851ff79ac Mon Sep 17 00:00:00 2001
From: cccclai <chenlai@meta.com>
Date: Mon, 13 Oct 2025 19:18:21 -0700
Subject: [PATCH] Document Quantizer API and precision options for MTK backend

Added Quantizer API section with precision configuration example.
---
 docs/source/backends-mediatek.md | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/docs/source/backends-mediatek.md b/docs/source/backends-mediatek.md
index 34cd56f971b..cbd697440c4 100644
--- a/docs/source/backends-mediatek.md
+++ b/docs/source/backends-mediatek.md
@@ -38,6 +38,17 @@ cd executorch
 ```
 The exported `.pte` file is saved in a directory named after the model.
 
+### Quantizer API
+
+Quantizer can be configured with different precision. We currently support A16W16, A16W8, A16W4, A8W8 and A8W4
+
+The example code will be
+```python
+precision = "A16W16"
+quantizer = NeuropilotQuantizer()
+quantizer.setup_precision(getattr(Precision, precision))
+```
+
 ### Partitioner API
 
 A list of CompileSpec is suppported by MediaTek backend: