Add support for OCI OpenAI Responses API with Langchain/Langgraph

libs/oci/README.md

@@ -79,6 +79,53 @@ structured_llm = llm.with_structured_output(Joke)
 structured_llm.invoke("Tell me a joke about programming")
 ```
 
+### 5. Use OpenAI Responses API
+`ChatOCIOpenAI` supports OpenAI Responses API.
+
+```python
+from oci_openai import (
+    OciSessionAuth,
+)
+from langchain_oci import ChatOCIOpenAI
+client = ChatOCIOpenAI(
+        auth=OciSessionAuth(profile_name="MY_PROFILE_NAME"),
+        compartment_id="MY_COMPARTMENT_ID",
+        region="us-chicago-1",
+        model="openai.gpt-4.1",
+        conversation_store_id="MY_CONVERSATION_STORE_ID"
+    )
+messages = [
+        (
+            "system",
+            "You are a helpful translator. Translate the user sentence to French.",
+        ),
+        ("human", "I love programming."),
+    ]
+response = client.invoke(messages)
+```
+NOTE: By default `store` argument is set to `True` which requires passing `conversation_store_id`. You can set `store` to `False` and not pass `conversation_store_id`.
+```python
+from oci_openai import (
+    OciSessionAuth,
+)
+from langchain_oci import ChatOCIOpenAI
+client = ChatOCIOpenAI(
+        auth=OciSessionAuth(profile_name="MY_PROFILE_NAME"),
+        compartment_id="MY_COMPARTMENT_ID",
+        region="us-chicago-1",
+        model="openai.gpt-4.1",
+        store=False
+    )
+messages = [
+        (
+            "system",
+            "You are a helpful translator. Translate the user sentence to French.",
+        ),
+        ("human", "I love programming."),
+    ]
+response = client.invoke(messages)
+```
+
 
 ## OCI Data Science Model Deployment Examples
 

libs/oci/examples/chat_models/langchain_oci_openai_responses_api_example.py

@@ -0,0 +1,118 @@
+from langchain_core.prompts import ChatPromptTemplate
+from oci_openai import OciSessionAuth
+from pydantic import BaseModel, Field
+from rich import print
+
+from langchain_oci import ChatOCIOpenAI
+
+COMPARTMENT_ID = "ocid1.compartment.oc1..aaaaaaaaexample"
+CONVERSATION_STORE_ID = (
+    "ocid1.generativeaiconversationstore.oc1.us-chicago-1.aaaaaaaaexample"
+)
+SERVICE_ENDPOINT = "https://inference.generativeai.us-chicago-1.oci.oraclecloud.com"
+REGION = "us-chicago-1"
+MODEL = "openai.gpt-4o"
+PROFILE_NAME = "oc1"
+
+
+def get_oci_openai_client():
+    return ChatOCIOpenAI(
+        auth=OciSessionAuth(profile_name=PROFILE_NAME),
+        compartment_id=COMPARTMENT_ID,
+        service_endpoint=SERVICE_ENDPOINT,
+        model=MODEL,
+        conversation_store_id=CONVERSATION_STORE_ID,
+    )
+
+
+def do_model_invoke():
+    client = get_oci_openai_client()
+    messages = [
+        (
+            "system",
+            "You are a helpful translator. Translate the user sentence to French.",
+        ),
+        ("human", "I love programming."),
+    ]
+    response = client.invoke(messages)
+    return response
+
+
+def do_prompt_chaining():
+    client = get_oci_openai_client()
+    prompt = ChatPromptTemplate.from_messages(
+        [
+            (
+                "system",
+                "You are a helpful assistant that translates {input_language}"
+                " to {output_language}.",
+            ),
+            ("human", "{input}"),
+        ]
+    )
+
+    chain = prompt | client
+    response = chain.invoke(
+        {
+            "input_language": "English",
+            "output_language": "German",
+            "input": "I love programming.",
+        }
+    )
+    return response
+
+
+def do_function_calling():
+    class GetWeather(BaseModel):
+        """Get the current weather in a given location"""
+
+        location: str = Field(
+            ..., description="The city and state, e.g. San Francisco, CA"
+        )
+
+    client = get_oci_openai_client()
+    llm_with_tools = client.bind_tools([GetWeather])
+    response = llm_with_tools.invoke(
+        "what is the weather like in San Francisco",
+    )
+    return response
+
+
+def do_web_search():
+    client = get_oci_openai_client()
+    tool = {"type": "web_search"}
+    llm_with_tools = client.bind_tools([tool])
+
+    response = llm_with_tools.invoke("What was a positive news story from today?")
+    return response
+
+
+def do_hosted_mcp_calling():
+    client = get_oci_openai_client()
+    llm_with_mcp_tools = client.bind_tools(
+        [
+            {
+                "type": "mcp",
+                "server_label": "deepwiki",
+                "server_url": "https://mcp.deepwiki.com/mcp",
+                "require_approval": "never",
+            }
+        ]
+    )
+    response = llm_with_mcp_tools.invoke(
+        "What transport protocols does the 2025-03-26 version of the MCP "
+        "spec (modelcontextprotocol/modelcontextprotocol) support?"
+    )
+    return response
+
+
+def main():
+    print(do_model_invoke())
+    print(do_prompt_chaining())
+    print(do_function_calling())
+    print(do_web_search())
+    print(do_hosted_mcp_calling())
+
+
+if __name__ == "__main__":
+    main()

libs/oci/langchain_oci/__init__.py

@@ -6,7 +6,7 @@
     ChatOCIModelDeploymentTGI,
     ChatOCIModelDeploymentVLLM,
 )
-from langchain_oci.chat_models.oci_generative_ai import ChatOCIGenAI
+from langchain_oci.chat_models.oci_generative_ai import ChatOCIGenAI, ChatOCIOpenAI
 from langchain_oci.embeddings.oci_data_science_model_deployment_endpoint import (
     OCIModelDeploymentEndpointEmbeddings,
 )
@@ -32,4 +32,5 @@
     "OCIModelDeploymentLLM",
     "OCIModelDeploymentTGI",
     "OCIModelDeploymentVLLM",
+    "ChatOCIOpenAI",
 ]