Chat Models¶

UiPath provides chat model classes for several providers (OpenAI via UiPathAzureChatOpenAI, Anthropic on AWS Bedrock via UiPathChatAnthropicBedrock, Google Vertex AI via UiPathChatGoogleGenerativeAI, and more), plus the generic UiPathChat. These are compatible with LangGraph as drop in replacements. You do not need to add tokens from OpenAI, Anthropic, or Google, usage of these chat models will consume Agent Units on your account.

Available models¶

LLM models are served through the UiPath LLM Gateway and are subject to AI Trust Layer policies, so the exact set of models available to you depends on your tenant configuration. List the models you can use with the uipath CLI:

$ uipath list-models
                                   Available LLM Models
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━┓
┃ AwsBedrock                               ┃ OpenAi                  ┃ VertexAi         ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━┩
│ anthropic.claude-haiku-4-5-20251001-v1:0 │ gpt-4.1-2025-04-14      │ gemini-2.5-flash │
│ anthropic.claude-opus-4-7                │ gpt-4.1-mini-2025-04-14 │ gemini-2.5-pro   │
│ ...                                      │ ...                     │ ...              │
└──────────────────────────────────────────┴─────────────────────────┴──────────────────┘

Pick a model id from the relevant provider column and pass it to the matching chat model class:

from uipath_langchain.chat import (
    UiPathAzureChatOpenAI,
    UiPathChatAnthropicBedrock,
    UiPathChatGoogleGenerativeAI,
)

# AWS Bedrock (Anthropic) models
llm = UiPathChatAnthropicBedrock(model="anthropic.claude-haiku-4-5-20251001-v1:0")

# OpenAI models
llm = UiPathAzureChatOpenAI(model="gpt-4.1-mini-2025-04-14")

# Google Vertex AI models
llm = UiPathChatGoogleGenerativeAI(model="gemini-2.5-flash")

Passthrough vs normalized

The provider-specific classes (UiPathAzureChatOpenAI, UiPathChatAnthropicBedrock, UiPathChatGoogleGenerativeAI) are passthrough: requests and responses go directly to the provider API, so you get the provider's full, native feature set. UiPathChat is normalized: it exposes a single unified interface across providers at the cost of provider-specific capabilities. We recommend the passthrough classes.

UiPathAzureChatOpenAI¶

UiPathAzureChatOpenAI can be used as a drop in replacement for ChatOpenAI or AzureChatOpenAI.

Example usage¶

Here is a code that is using ChatOpenAI

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="gpt-4o",
    temperature=0,
    max_tokens=4000,
    timeout=30,
    max_retries=2,
    # api_key="...",  # if you prefer to pass api key in directly instead of using env vars
    # base_url="...",
    # organization="...",
    # other params...
)

You can simply change ChatOpenAi with UiPathAzureChatOpenAI, you don't have to provide an OpenAI token.

from uipath_langchain.chat.models import UiPathAzureChatOpenAI

llm = UiPathAzureChatOpenAI(
    model="gpt-4.1-mini-2025-04-14",
    temperature=0,
    max_tokens=4000,
    timeout=30,
    max_retries=2,
    # other params...
)

UiPathAzureChatOpenAI supports the OpenAI models (the OpenAi column of uipath list-models).

UiPathChat¶

UiPathChat is a more versatile class that can suport models from diferent vendors including OpenAI.

Example usage¶

Given the following code:

from langchain_anthropic import ChatAnthropic

llm = ChatAnthropic(
    model="claude-3-5-sonnet-20240620",
    temperature=0,
    max_tokens=1024,
    timeout=None,
    max_retries=2,
    # other params...
)

You can replace it with UiPathChat like so:

from uipath_langchain.chat.models import UiPathChat

llm = UiPathChat(
    model="anthropic.claude-3-opus-20240229-v1:0",
    temperature=0,
    max_tokens=1024,
    timeout=None,
    max_retries=2,
    # other params...
)

UiPathChat supports models from multiple vendors (AWS Bedrock, OpenAI, Google Vertex AI), so it can use any model from uipath list-models.

Warning

Please note that you may get errors related to data residency, as some models are not available on all regions.

Example: [Enforced Region] No model configuration found for product uipath-python-sdk in EU using model anthropic.claude-3-opus-20240229-v1:0.