feat: retry HTTP 408/502/503/504 by default and default max_retries to 3 (#89)

ionut-mihalache-uipath · claude · web-flow · commit 3c8ebcc8cf6f · 2026-05-28T13:10:23.000+03:00
Co-authored-by: Claude Opus 4.7 (1M context) &lt;noreply@anthropic.com&gt;
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -2,6 +2,15 @@
 
 All notable changes to `uipath_llm_client` (core package) will be documented in this file.
 
+## [1.13.0] - 2026-05-27
+
+### Added
+- `UiPathRequestTimeoutError` (HTTP 408) and `UiPathBadGatewayError` (HTTP 502) exception classes. Both are registered in `_STATUS_CODE_TO_EXCEPTION`, re-exported from `uipath.llm_client`, and inherit from `UiPathAPIError` for compatibility with existing handlers.
+
+### Changed
+- **Default retry set expanded.** `_DEFAULT_RETRY_ON_EXCEPTIONS` in `uipath.llm_client.utils.retry` now covers `UiPathRequestTimeoutError` (408), `UiPathRateLimitError` (429), `UiPathBadGatewayError` (502), `UiPathServiceUnavailableError` (503), `UiPathGatewayTimeoutError` (504), and `UiPathTooManyRequestsError` (529) — up from `{429, 529}`. Applies to every provider client (`UiPathOpenAI`, `UiPathAnthropic*`, `UiPathGoogle`) since they all share the same `UiPathHttpxClient`-backed retry transport. `Retry-After` / `x-retry-after` headers and exponential backoff with jitter behave as before.
+- **`UiPathHttpxClient` / `UiPathHttpxAsyncClient` default `max_retries` raised from `0` to `3`.** Callers that pass `max_retries=None` (or omit it entirely) now get 3 retries by default. Pass `max_retries=0` explicitly to opt out — `max_retries=0` continues to disable retries, so the opt-out path is unchanged.
+
 ## [1.12.2] - 2026-05-24
 
 ### Changed
diff --git a/packages/uipath_langchain_client/CHANGELOG.md b/packages/uipath_langchain_client/CHANGELOG.md
@@ -2,6 +2,12 @@
 
 All notable changes to `uipath_langchain_client` will be documented in this file.
 
+## [1.13.0] - 2026-05-27
+
+### Changed
+- **`UiPathBaseLLMClient.max_retries` field default raised from `0` to `3`.** Every LangChain chat and embedding client built on this base (`UiPathChat`, `UiPathChatOpenAI`, `UiPathAzureChatOpenAI`, `UiPathChatAnthropic`, `UiPathChatAnthropicBedrock`, `UiPathChatBedrock`, `UiPathChatBedrockConverse`, `UiPathChatVertexAI`, `UiPathChatFireworks`, `UiPathChatLiteLLM`, plus the matching embeddings classes) now retries failed requests 3 times by default. Pass `max_retries=0` explicitly to disable retries — the opt-out path is unchanged. Combined with the expanded default retry set in `uipath-llm-client` 1.13.0, every LangChain client now retries on HTTP 408, 429, 502, 503, 504, and 529 out of the box.
+- Bumped `uipath-llm-client` floor to `>=1.13.0` to pick up the expanded default retry set and the new `UiPathRequestTimeoutError` / `UiPathBadGatewayError` typed exceptions.
+
 ## [1.12.2] - 2026-05-24
 
 ### Changed
diff --git a/packages/uipath_langchain_client/pyproject.toml b/packages/uipath_langchain_client/pyproject.toml
@@ -6,7 +6,7 @@ readme = "README.md"
 requires-python = ">=3.11"
 dependencies = [
     "langchain>=1.2.15,<2.0.0",
-    "uipath-llm-client>=1.12.2,<2.0.0",
+    "uipath-llm-client>=1.13.0,<2.0.0",
 ]
 
 [project.optional-dependencies]
diff --git a/packages/uipath_langchain_client/src/uipath_langchain_client/__version__.py b/packages/uipath_langchain_client/src/uipath_langchain_client/__version__.py
@@ -1,3 +1,3 @@
 __title__ = "UiPath LangChain Client"
 __description__ = "A Python client for interacting with UiPath's LLM services via LangChain."
-__version__ = "1.12.2"
+__version__ = "1.13.0"
diff --git a/packages/uipath_langchain_client/src/uipath_langchain_client/base_client.py b/packages/uipath_langchain_client/src/uipath_langchain_client/base_client.py
@@ -145,8 +145,8 @@ class UiPathBaseLLMClient(BaseModel, ABC):
         description="Client-side request timeout in seconds",
     )
     max_retries: int = Field(
-        default=0,
-        description="Maximum number of retries for failed requests",
+        default=3,
+        description="Maximum number of retries for failed requests. Pass 0 to disable retries.",
     )
     retry_config: RetryConfig | None = Field(
         default=None,
diff --git a/src/uipath/llm_client/__init__.py b/src/uipath/llm_client/__init__.py
@@ -39,13 +39,15 @@
 from uipath.llm_client.utils.exceptions import (
     UiPathAPIError,
     UiPathAuthenticationError,
+    UiPathBadGatewayError,
     UiPathBadRequestError,
     UiPathConflictError,
     UiPathGatewayTimeoutError,
     UiPathInternalServerError,
     UiPathNotFoundError,
     UiPathPermissionDeniedError,
     UiPathRateLimitError,
+    UiPathRequestTimeoutError,
     UiPathRequestTooLargeError,
     UiPathServiceUnavailableError,
     UiPathTooManyRequestsError,
@@ -69,13 +71,15 @@
     # Exceptions
     "UiPathAPIError",
     "UiPathAuthenticationError",
+    "UiPathBadGatewayError",
     "UiPathBadRequestError",
     "UiPathConflictError",
     "UiPathGatewayTimeoutError",
     "UiPathInternalServerError",
     "UiPathNotFoundError",
     "UiPathPermissionDeniedError",
     "UiPathRateLimitError",
+    "UiPathRequestTimeoutError",
     "UiPathRequestTooLargeError",
     "UiPathServiceUnavailableError",
     "UiPathTooManyRequestsError",
diff --git a/src/uipath/llm_client/__version__.py b/src/uipath/llm_client/__version__.py
@@ -1,3 +1,3 @@
 __title__ = "UiPath LLM Client"
 __description__ = "A Python client for interacting with UiPath's LLM services."
-__version__ = "1.12.2"
+__version__ = "1.13.0"
diff --git a/src/uipath/llm_client/httpx_client.py b/src/uipath/llm_client/httpx_client.py
@@ -68,6 +68,10 @@
 # Sentinel to distinguish "not provided" from an explicit ``None`` / ``False``.
 _UNSET: Any = object()
 
+# Default applied when ``max_retries`` is left as ``None``. Callers can still
+# opt out by passing ``max_retries=0`` explicitly.
+_DEFAULT_MAX_RETRIES: typing.Final[int] = 3
+
 
 class UiPathHttpxClient(Client):
     """Synchronous HTTP client configured for UiPath LLM services.
@@ -135,8 +139,8 @@ def __init__(
             captured_headers: Case-insensitive header name prefixes to capture from
                 responses. Captured headers are stored in a ContextVar and can be
                 retrieved with get_captured_response_headers(). Defaults to ("x-uipath-",).
-            max_retries: Maximum retry attempts for failed requests. Defaults to 0
-                (retries disabled). Set to a positive integer to enable retries.
+            max_retries: Maximum retry attempts for failed requests. Defaults to 3
+                when left as ``None``. Pass ``0`` to disable retries explicitly.
             retry_config: Custom retry configuration (backoff, retryable status codes).
             logger: Logger instance for request/response logging.
             auth: HTTP authentication (same as httpx.Client).  Derived from
@@ -193,7 +197,7 @@ def __init__(
         # Setup retry transport if not provided
         if transport is None:
             transport = RetryableHTTPTransport(
-                max_retries=max_retries if max_retries is not None else 0,
+                max_retries=max_retries if max_retries is not None else _DEFAULT_MAX_RETRIES,
                 retry_config=retry_config,
                 logger=logger,
             )
@@ -354,7 +358,7 @@ def __init__(
         # Setup retry transport if not provided
         if transport is None:
             transport = RetryableAsyncHTTPTransport(
-                max_retries=max_retries if max_retries is not None else 0,
+                max_retries=max_retries if max_retries is not None else _DEFAULT_MAX_RETRIES,
                 retry_config=retry_config,
                 logger=logger,
             )
diff --git a/src/uipath/llm_client/utils/exceptions.py b/src/uipath/llm_client/utils/exceptions.py
@@ -118,6 +118,12 @@ class UiPathNotFoundError(UiPathAPIError):
     status_code: Literal[404] = 404  # pyright: ignore[reportIncompatibleVariableOverride]
 
 
+class UiPathRequestTimeoutError(UiPathAPIError):
+    """HTTP 408 Request Timeout error."""
+
+    status_code: Literal[408] = 408  # pyright: ignore[reportIncompatibleVariableOverride]
+
+
 class UiPathConflictError(UiPathAPIError):
     """HTTP 409 Conflict error."""
 
@@ -211,6 +217,12 @@ class UiPathInternalServerError(UiPathAPIError):
     status_code: Literal[500] = 500  # pyright: ignore[reportIncompatibleVariableOverride]
 
 
+class UiPathBadGatewayError(UiPathAPIError):
+    """HTTP 502 Bad Gateway error."""
+
+    status_code: Literal[502] = 502  # pyright: ignore[reportIncompatibleVariableOverride]
+
+
 class UiPathServiceUnavailableError(UiPathAPIError):
     """HTTP 503 Service Unavailable error."""
 
@@ -234,11 +246,13 @@ class UiPathTooManyRequestsError(UiPathAPIError):
     401: UiPathAuthenticationError,
     403: UiPathPermissionDeniedError,
     404: UiPathNotFoundError,
+    408: UiPathRequestTimeoutError,
     409: UiPathConflictError,
     413: UiPathRequestTooLargeError,
     422: UiPathUnprocessableEntityError,
     429: UiPathRateLimitError,
     500: UiPathInternalServerError,
+    502: UiPathBadGatewayError,
     503: UiPathServiceUnavailableError,
     504: UiPathGatewayTimeoutError,
     529: UiPathTooManyRequestsError,
@@ -266,11 +280,13 @@ def raise_for_status() -> Response:
     "UiPathAuthenticationError",
     "UiPathPermissionDeniedError",
     "UiPathNotFoundError",
+    "UiPathRequestTimeoutError",
     "UiPathConflictError",
     "UiPathRequestTooLargeError",
     "UiPathUnprocessableEntityError",
     "UiPathRateLimitError",
     "UiPathInternalServerError",
+    "UiPathBadGatewayError",
     "UiPathServiceUnavailableError",
     "UiPathGatewayTimeoutError",
     "UiPathTooManyRequestsError",
diff --git a/src/uipath/llm_client/utils/retry.py b/src/uipath/llm_client/utils/retry.py
@@ -49,13 +49,22 @@
 
 from uipath.llm_client.utils.exceptions import (
     UiPathAPIError,
+    UiPathBadGatewayError,
+    UiPathGatewayTimeoutError,
     UiPathRateLimitError,
+    UiPathRequestTimeoutError,
+    UiPathServiceUnavailableError,
     UiPathTooManyRequestsError,
 )
 
 # Default retry configuration values
+# Status codes retried by default: 408, 429, 502, 503, 504, 529.
 _DEFAULT_RETRY_ON_EXCEPTIONS: tuple[type[Exception], ...] = (
+    UiPathRequestTimeoutError,
     UiPathRateLimitError,
+    UiPathBadGatewayError,
+    UiPathServiceUnavailableError,
+    UiPathGatewayTimeoutError,
     UiPathTooManyRequestsError,
 )
 _DEFAULT_INITIAL_DELAY: float = 2.0
@@ -127,7 +136,7 @@ class RetryConfig(TypedDict):
 
     Attributes:
         retry_on_exceptions: Tuple of exception types to retry on.
-            Defaults to (UiPathRateLimitError,).
+            Defaults to the typed exceptions for HTTP 408, 429, 502, 503, 504, 529.
         initial_delay: Initial delay in seconds before first retry.
             Defaults to 2.0.
         max_delay: Maximum delay in seconds between retries.
diff --git a/tests/core/features/test_exceptions.py b/tests/core/features/test_exceptions.py
@@ -9,12 +9,14 @@
 from uipath.llm_client.utils.exceptions import (
     UiPathAPIError,
     UiPathAuthenticationError,
+    UiPathBadGatewayError,
     UiPathBadRequestError,
     UiPathGatewayTimeoutError,
     UiPathInternalServerError,
     UiPathNotFoundError,
     UiPathPermissionDeniedError,
     UiPathRateLimitError,
+    UiPathRequestTimeoutError,
     UiPathServiceUnavailableError,
     UiPathTooManyRequestsError,
     patch_raise_for_status,
@@ -170,8 +172,10 @@ def test_all_status_code_mappings(self):
             401: UiPathAuthenticationError,
             403: UiPathPermissionDeniedError,
             404: UiPathNotFoundError,
+            408: UiPathRequestTimeoutError,
             429: UiPathRateLimitError,
             500: UiPathInternalServerError,
+            502: UiPathBadGatewayError,
             503: UiPathServiceUnavailableError,
             504: UiPathGatewayTimeoutError,
             529: UiPathTooManyRequestsError,
diff --git a/tests/core/features/test_httpx_client.py b/tests/core/features/test_httpx_client.py
@@ -80,6 +80,24 @@ def test_client_with_retry_config(self):
         assert isinstance(client._transport, RetryableHTTPTransport)
         client.close()
 
+    def test_client_default_max_retries_is_three(self):
+        """Caller passing no ``max_retries`` should get the 3-retry default."""
+        from uipath.llm_client.httpx_client import UiPathHttpxClient
+
+        client = UiPathHttpxClient(base_url="https://example.com")
+        assert isinstance(client._transport, RetryableHTTPTransport)
+        assert client._transport.retryer is not None
+        client.close()
+
+    def test_client_explicit_zero_disables_retries(self):
+        """Passing ``max_retries=0`` must still disable retries."""
+        from uipath.llm_client.httpx_client import UiPathHttpxClient
+
+        client = UiPathHttpxClient(base_url="https://example.com", max_retries=0)
+        assert isinstance(client._transport, RetryableHTTPTransport)
+        assert client._transport.retryer is None
+        client.close()
+
     def test_client_with_byo_connection_id(self):
         """Test client adds BYO connection ID header."""
         from uipath.llm_client.httpx_client import UiPathHttpxClient
@@ -123,6 +141,22 @@ def test_async_client_with_retry_config(self):
         # Transport should be RetryableAsyncHTTPTransport
         assert isinstance(client._transport, RetryableAsyncHTTPTransport)
 
+    def test_async_client_default_max_retries_is_three(self):
+        """Async caller passing no ``max_retries`` should get the 3-retry default."""
+        from uipath.llm_client.httpx_client import UiPathHttpxAsyncClient
+
+        client = UiPathHttpxAsyncClient(base_url="https://example.com")
+        assert isinstance(client._transport, RetryableAsyncHTTPTransport)
+        assert client._transport.retryer is not None
+
+    def test_async_client_explicit_zero_disables_retries(self):
+        """Async client: passing ``max_retries=0`` must still disable retries."""
+        from uipath.llm_client.httpx_client import UiPathHttpxAsyncClient
+
+        client = UiPathHttpxAsyncClient(base_url="https://example.com", max_retries=0)
+        assert isinstance(client._transport, RetryableAsyncHTTPTransport)
+        assert client._transport.retryer is None
+
 
 class TestBuildRoutingHeaders:
     """Tests for build_routing_headers function."""
diff --git a/tests/core/features/test_retry.py b/tests/core/features/test_retry.py
diff --git a/tests/core/smoke_test.py b/tests/core/smoke_test.py
diff --git a/tests/langchain/features/test_default_max_retries.py b/tests/langchain/features/test_default_max_retries.py

Original file line number	Diff line number	Diff line change
`@@ -6,7 +6,7 @@ readme = "README.md"`
`6`	`6`	`requires-python = ">=3.11"`
`7`	`7`	`dependencies = [`
`8`	`8`	`"langchain>=1.2.15,<2.0.0",`
`9`		`- "uipath-llm-client>=1.12.2,<2.0.0",`
	`9`	`+ "uipath-llm-client>=1.13.0,<2.0.0",`
`10`	`10`	`]`
`11`	`11`
`12`	`12`	`[project.optional-dependencies]`
Original file line number	Diff line number	Diff line change
`@@ -145,8 +145,8 @@ class UiPathBaseLLMClient(BaseModel, ABC):`
`145`	`145`	`description="Client-side request timeout in seconds",`
`146`	`146`	`)`
`147`	`147`	`max_retries: int = Field(`
`148`		`- default=0,`
`149`		`- description="Maximum number of retries for failed requests",`
	`148`	`+ default=3,`
	`149`	`+ description="Maximum number of retries for failed requests. Pass 0 to disable retries.",`
`150`	`150`	`)`
`151`	`151`	`retry_config: RetryConfig \| None = Field(`
`152`	`152`	`default=None,`