diff --git a/CHANGELOG.md b/CHANGELOG.md
index c031bcc..78f55e9 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,5 +1,11 @@
 # Changelog
 
+## [2.0.3](https://github.com/ScrapingBee/scrapingbee-python/compare/v2.0.2...v2.0.3) (2026-06-30)
+
+### Improvement
+
+- Document Auto-Mode (`mode=auto`) support: ScrapingBee picks the cheapest scraping configuration that succeeds and charges only for the winning one. Read the credits charged from the `Spb-auto-cost` response header, and optionally cap the cost with `max_cost`. No client changes are required — these are pass-through query parameters.
+
 ## [2.0.0](https://github.com/ScrapingBee/scrapingbee-python/compare/v1.2.0...v2.0.0) (2023-10-03)
 
 ### Improvement
diff --git a/README.md b/README.md
index 328dc4d..4ed1691 100644
--- a/README.md
+++ b/README.md
@@ -95,6 +95,36 @@ You can find all the supported parameters on [ScrapingBee's documentation](https
 
 You can send custom cookies and headers like you would normally do with the requests library.
 
+## Auto-Mode
+
+With Auto-Mode, ScrapingBee picks the cheapest scraping configuration that successfully scrapes the page for you: it tries the cheaper options first and stops at the first one that works. You are charged only for the winning configuration (and 0 credits if every configuration fails).
+
+```python
+>>> from scrapingbee import ScrapingBeeClient
+
+>>> client = ScrapingBeeClient(api_key='REPLACE-WITH-YOUR-API-KEY')
+
+# Auto-Mode: ScrapingBee picks the cheapest config that works; you're charged only for the winning one.
+>>> response = client.get(
+    'https://example.com',
+    params={
+        'mode': 'auto',
+        # Optional: cap the credits a single request may cost (omit for uncapped).
+        'max_cost': 25
+    }
+)
+
+# Spb-auto-cost reports the credits actually charged (0 if every config failed).
+>>> response.headers['Spb-auto-cost']
+'1'
+```
+
+Notes:
+
+- Auto-Mode is only available on `GET` requests.
+- `max_cost` is optional and must be `>= 1`; omit it to leave the cost uncapped.
+- `mode=auto` cannot be combined with `render_js`, `premium_proxy`, or `stealth_proxy` (ScrapingBee chooses these for you). Sending them together returns a `400`.
+
 ## Screenshot
 
 Here a little exemple on how to retrieve and store a screenshot from the ScrapingBee blog in its mobile resolution.
diff --git a/scrapingbee/__version__.py b/scrapingbee/__version__.py
index 0309ae2..5fa9130 100644
--- a/scrapingbee/__version__.py
+++ b/scrapingbee/__version__.py
@@ -1 +1 @@
-__version__ = "2.0.2"
+__version__ = "2.0.3"
diff --git a/tests/test_utils.py b/tests/test_utils.py
index 583e497..4b96807 100644
--- a/tests/test_utils.py
+++ b/tests/test_utils.py
@@ -18,7 +18,7 @@ def test_process_headers():
     """It should add a Spb- prefix to header names"""
     output = process_headers({"Accept-Language": "En-US"})
     assert output == {
-        "User-Agent": "ScrapingBee-Python/2.0.2",
+        "User-Agent": "ScrapingBee-Python/2.0.3",
         "Spb-Accept-Language": "En-US",
     }