UiPath · cosminacho · Apr 8, 2026 · Apr 4, 2026 · Apr 4, 2026 · Apr 5, 2026
@@ -2,6 +2,23 @@
 
 All notable changes to `uipath_llm_client` (core package) will be documented in this file.
 
+## [1.8.0] - 2026-04-08
+
+### Added
+- `UiPathLiteLLM` — provider-agnostic LLM client powered by LiteLLM
+  - `completion` / `acompletion` for chat completions across all providers
+  - `embedding` / `aembedding` for embeddings
+  - Automatic model discovery from the UiPath backend — detects vendor, api_flavor, and model family
+  - Optional `vendor_type` and `api_flavor` overrides (same pattern as LangChain factory)
+  - Supports OpenAI (chat-completions + responses API), Gemini, Bedrock (invoke + converse), and Vertex AI Claude
+  - All HTTP routed through UiPath httpx transport (auth, retry, headers) — no direct calls to Google/AWS/OpenAI
+  - Explicit completion parameters with full IDE autocomplete
+- `litellm` as an optional dependency (`uv add uipath-llm-client[litellm]`)
+- `_strict_response_validation` parameter to all Anthropic client classes
+
+### Changed
+- Updated dependency versions: `uipath-platform>=0.1.21`, `anthropic>=0.91.0`, `litellm>=1.83.4`
+
 ## [1.7.0] - 2026-04-03
 
 ### Added

@@ -1,14 +1,13 @@
 # UiPath LLM Client
 
-A Python client for interacting with UiPath's LLM services. This package provides both a low-level HTTP client and framework-specific integrations (LangChain, LlamaIndex) for accessing LLMs through UiPath's infrastructure.
+A Python client for interacting with UiPath's LLM services. This package provides both a low-level HTTP client and a LangChain integration for accessing LLMs through UiPath's infrastructure.
 
 ## Architecture Overview
 
 This repository is organized as a monorepo with the following packages:
 
 - **`uipath_llm_client`** (root): Core HTTP client with authentication, retry logic, and request handling
 - **`uipath_langchain_client`** (packages/): LangChain-compatible chat models and embeddings
-- **`uipath_llamaindex_client`** (packages/): LlamaIndex-compatible integrations
 
 ### Supported Backends
 
@@ -816,7 +815,7 @@ uv run pyright
 
 ### Testing
 
-Tests use [VCR.py](https://vcrpy.readthedocs.io/) to record and replay HTTP interactions. Cassettes (recorded responses) are stored in `tests/cassettes/` using Git LFS.
+Tests use [VCR.py](https://vcrpy.readthedocs.io/) to record and replay HTTP interactions. Cassettes (recorded responses) are stored in `tests/cassettes.db` (SQLite) via `pytest-recording`.
 
 **Important:** Tests must pass locally before submitting a PR. The CI pipeline does not make any real API requests—it only runs tests using the pre-recorded cassettes.
 
@@ -855,7 +854,9 @@ uipath-llm-client/
 │   ├── clients/                        # Native SDK wrappers
 │   │   ├── openai/                     # UiPathOpenAI, UiPathAzureOpenAI, etc.
 │   │   ├── anthropic/                  # UiPathAnthropic, UiPathAnthropicBedrock, etc.
-│   │   └── google/                     # UiPathGoogle
+│   │   ├── google/                     # UiPathGoogle
+│   │   ├── normalized/                 # UiPathNormalizedClient (provider-agnostic)
+│   │   └── litellm/                    # UiPathLiteLLM (via LiteLLM)
 │   ├── settings/                       # Backend-specific settings & auth
 │   │   ├── base.py                     # UiPathBaseSettings, UiPathAPIConfig
 │   │   ├── platform/                   # PlatformSettings, PlatformAuth
@@ -878,9 +879,9 @@ uipath-llm-client/
 │   │           ├── vertexai/           # UiPathChatAnthropicVertex
 │   │           ├── bedrock/            # UiPathChatBedrock, UiPathChatBedrockConverse
 │   │           ├── fireworks/          # UiPathChatFireworks, UiPathFireworksEmbeddings
+│   │           ├── litellm/           # UiPathChatLiteLLM, UiPathLiteLLMEmbeddings
 │   │           └── azure/              # UiPathAzureAIChatCompletionsModel
-│   └── uipath_llamaindex_client/       # LlamaIndex integration (planned)
-└── tests/                              # Test suite with VCR cassettes
+└── tests/                              # Test suite with VCR cassettes (SQLite)
 ```
 
 ## License

@@ -2,6 +2,16 @@
 
 All notable changes to `uipath_langchain_client` will be documented in this file.
 
+## [1.8.0] - 2026-04-08
+
+### Added
+- `UiPathChatLiteLLM` — LangChain chat model powered by LiteLLM, supporting all UiPath gateway providers
+- `langchain-litellm` as an optional dependency for LiteLLM integration
+
+### Changed
+- Updated dependency versions: `anthropic[bedrock,vertex]>=0.91.0`
+- Version bump to match core package 1.8.0
+
 ## [1.7.1] - 2026-04-04
 
 ### Added

@@ -5,35 +5,38 @@ dynamic = ["version"]
 readme = "README.md"
 requires-python = ">=3.11"
 dependencies = [
-    "langchain>=1.2.13",
-    "uipath-llm-client>=1.7.0",
+    "langchain>=1.2.15",
+    "uipath-llm-client>=1.8.0",
 ]
 
 [project.optional-dependencies]
 openai = [
-    "langchain-openai>=1.1.11",
+    "langchain-openai>=1.1.12",
 ]
 google = [
     "langchain-google-genai>=4.2.1",
 ]
 anthropic = [
     "langchain-anthropic>=1.4.0",
-    "anthropic[bedrock,vertex]>=0.86.0",
+    "anthropic[bedrock,vertex]>=0.91.0",
 ]
 aws = [
-    "langchain-aws[anthropic]>=1.4.1",
+    "langchain-aws[anthropic]>=1.4.3",
 ]
 vertexai = [
     "langchain-google-vertexai>=3.2.2",
 ]
 azure = [
-    "langchain-azure-ai>=1.1.1",
+    "langchain-azure-ai>=1.2.1",
 ]
 fireworks = [
     "langchain-fireworks>=1.1.0",
 ]
+litellm = [
+    "langchain-litellm>=0.6.4",
+]
 all = [
-    "uipath-langchain-client[openai,aws,google,anthropic,azure,vertexai,fireworks]"
+    "uipath-langchain-client[openai,aws,google,anthropic,azure,vertexai,fireworks,litellm]"
 ]
 
 [build-system]

@@ -1,3 +1,3 @@
 __title__ = "UiPath LangChain Client"
 __description__ = "A Python client for interacting with UiPath's LLM services via LangChain."
-__version__ = "1.7.1"
+__version__ = "1.8.0"
@@ -1,3 +1,5 @@
+"""LangChain-compatible LLM client wrappers for UiPath providers."""
+
 from uipath_langchain_client.clients.normalized import UiPathChat, UiPathEmbeddings
 
 __all__ = [

@@ -1,3 +1,10 @@
+"""LangChain chat model for Anthropic Claude via UiPath.
+
+Requires the ``anthropic`` optional extra::
+
+    uv add uipath-langchain-client[anthropic]
+"""
+
 from uipath_langchain_client.clients.anthropic.chat_models import UiPathChatAnthropic
 
 __all__ = ["UiPathChatAnthropic"]
@@ -1,3 +1,10 @@
+"""LangChain chat models and embeddings for Azure AI via UiPath.
+
+Requires the ``azure`` optional extra::
+
+    uv add uipath-langchain-client[azure]
+"""
+
 from uipath_langchain_client.clients.azure.chat_models import UiPathAzureAIChatCompletionsModel
 from uipath_langchain_client.clients.azure.embeddings import UiPathAzureAIEmbeddingsModel
 

@@ -1,3 +1,10 @@
+"""LangChain chat models and embeddings for AWS Bedrock via UiPath.
+
+Requires the ``aws`` optional extra::
+
+    uv add uipath-langchain-client[aws]
+"""
+
 from uipath_langchain_client.clients.bedrock.chat_models import (
     UiPathChatAnthropicBedrock,
     UiPathChatBedrock,

@@ -1,3 +1,10 @@
+"""LangChain chat models and embeddings for Fireworks AI via UiPath.
+
+Requires the ``fireworks`` optional extra::
+
+    uv add uipath-langchain-client[fireworks]
+"""
+
 from uipath_langchain_client.clients.fireworks.chat_models import UiPathChatFireworks
 from uipath_langchain_client.clients.fireworks.embeddings import UiPathFireworksEmbeddings
 

@@ -1,3 +1,10 @@
+"""LangChain chat models and embeddings for Google Gemini via UiPath.
+
+Requires the ``google`` optional extra::
+
+    uv add uipath-langchain-client[google]
+"""
+
 from uipath_langchain_client.clients.google.chat_models import UiPathChatGoogleGenerativeAI
 from uipath_langchain_client.clients.google.embeddings import UiPathGoogleGenerativeAIEmbeddings
 

@@ -0,0 +1,11 @@
+"""UiPath LangChain LiteLLM chat model and embeddings.
+
+Requires the ``litellm`` optional extra::
+
+    uv add uipath-langchain-client[litellm]
+"""
+
+from uipath_langchain_client.clients.litellm.chat_models import UiPathChatLiteLLM
+from uipath_langchain_client.clients.litellm.embeddings import UiPathLiteLLMEmbeddings
+
+__all__ = ["UiPathChatLiteLLM", "UiPathLiteLLMEmbeddings"]
@@ -0,0 +1,144 @@
+# pyright: reportAttributeAccessIssue=false
+"""UiPath LangChain chat model powered by LiteLLM.
+
+Inherits ``UiPathBaseChatModel`` and ``ChatLiteLLM`` — UiPath handles
+authentication, routing, retries; LiteLLM handles provider-specific formatting.
+
+Example:
+    >>> from uipath_langchain_client.clients.litellm import UiPathChatLiteLLM
+    >>>
+    >>> chat = UiPathChatLiteLLM(model="gpt-5.2-2025-12-11")
+    >>> response = chat.invoke("Hello!")
+    >>> print(response.content)
+"""
+
+from typing import Any, Dict, Optional
+
+from langchain_core.callbacks import (
+    AsyncCallbackManagerForLLMRun,
+    CallbackManagerForLLMRun,
+)
+from pydantic import Field, model_validator
+from typing_extensions import Self
+
+from uipath.llm_client.clients.litellm import UiPathLiteLLM
+from uipath.llm_client.settings.constants import ApiFlavor, ApiType, RoutingMode, VendorType
+from uipath_langchain_client.base_client import UiPathBaseChatModel
+from uipath_langchain_client.settings import UiPathAPIConfig
+
+try:
+    from langchain_litellm import ChatLiteLLM
+except ImportError as e:
+    raise ImportError(
+        "The 'litellm' extra is required to use UiPathChatLiteLLM. "
+        "Install it with: uv add uipath-langchain-client[litellm]"
+    ) from e
+
+# Keys that the core UiPathLiteLLM handles internally — strip them
+# from ChatLiteLLM's kwargs before delegating to the core client.
+_CORE_HANDLED_KEYS = frozenset(
+    {
+        "model",
+        "api_base",
+        "api_key",
+        "custom_llm_provider",
+        "client",
+        "async_client",
+        "num_retries",
+        "max_retries",
+        "num_ctx",
+        "base_model",
+    }
+)
+
+
+class UiPathChatLiteLLM(UiPathBaseChatModel, ChatLiteLLM):  # type: ignore[override]
+    """LangChain chat model that routes through UiPath LLM Gateway via LiteLLM.
+
+    Discovers the model from the UiPath backend and uses LiteLLM for
+    provider-specific request formatting. Authentication, URL routing,
+    and retries are handled by the UiPath httpx client.
+
+    Args:
+        model: The model name (e.g., "gpt-5.2-2025-12-11", "gemini-2.5-flash").
+        settings: UiPath client settings. Defaults to environment-based settings.
+        vendor_type: Filter/override vendor type from discovery.
+        api_flavor: Override API flavor (e.g., ApiFlavor.RESPONSES, ApiFlavor.CONVERSE).
+    """
+
+    api_config: UiPathAPIConfig = Field(
+        default_factory=lambda: UiPathAPIConfig(
+            routing_mode=RoutingMode.PASSTHROUGH,
+            vendor_type=VendorType.OPENAI,  # placeholder — overridden by _setup_uipath
+            api_type=ApiType.COMPLETIONS,
+            freeze_base_url=True,
+        ),
+    )
+
+    vendor_type: VendorType | str | None = Field(default=None, exclude=True)
+    api_flavor: ApiFlavor | str | None = Field(default=None, exclude=True)
+
+    # Internal core client — handles discovery, HTTPHandler lifecycle, provider resolution
+    _core: UiPathLiteLLM | None = None
+
+    @model_validator(mode="after")
+    def _setup_uipath(self) -> Self:
+        """Create a UiPathLiteLLM core client and configure ChatLiteLLM fields."""
+        core = UiPathLiteLLM(
+            model_name=self.model_name,
+            byo_connection_id=self.byo_connection_id,
+            client_settings=self.client_settings,
+            vendor_type=self.vendor_type,
+            api_flavor=self.api_flavor,
+            timeout=self.request_timeout,
+            max_retries=self.max_retries,
+            default_headers=self.default_headers,
+            captured_headers=self.captured_headers,
+            retry_config=self.retry_config,
+            logger=self.logger,
+        )
+        self._core = core
+
+        # Sync api_config from discovery for UiPathBaseLLMClient httpx clients
+        self.api_config = UiPathAPIConfig(
+            vendor_type=core._api_config.vendor_type,
+            api_flavor=core._api_config.api_flavor,
+        )
+
+        # Set ChatLiteLLM fields so _default_params builds correct kwargs
+        self.custom_llm_provider = core._custom_llm_provider
+        self.api_key = "PLACEHOLDER"
+        self.api_base = str(core._completion_client.client.base_url)
+        self.model_name = core._litellm_model  # type: ignore[assignment]
+
+        return self
+
+    def completion_with_retry(
+        self, run_manager: Optional[CallbackManagerForLLMRun] = None, **kwargs: Any
+    ) -> Any:
+        """Delegate to the core UiPathLiteLLM.completion()."""
+        assert self._core is not None
+        for key in _CORE_HANDLED_KEYS:
+            kwargs.pop(key, None)
+        return self._core.completion(**kwargs)
+
+    async def acompletion_with_retry(
+        self, run_manager: Optional[AsyncCallbackManagerForLLMRun] = None, **kwargs: Any
+    ) -> Any:
+        """Delegate to the core UiPathLiteLLM.acompletion()."""
+        assert self._core is not None
+        for key in _CORE_HANDLED_KEYS:
+            kwargs.pop(key, None)
+        return await self._core.acompletion(**kwargs)
+
+    @property
+    def _llm_type(self) -> str:
+        return "uipath-litellm"
+
+    @property
+    def _identifying_params(self) -> Dict[str, Any]:
+        return {
+            "model": self.model_name,
+            "provider": self._core._custom_llm_provider if self._core else None,
+            "api_base": self.api_base,
+        }