Getting Started

export BASE_URL="https://your-deployment-url"
export AA_TOKEN="your-token"

from openai import OpenAI

client = OpenAI(
    base_url=f"{BASE_URL}/v1",
    api_key=AA_TOKEN,
)

import httpx
from pydantic_ai import Agent
from pydantic_ai.models.openai import OpenAIResponsesModel
from pydantic_ai.providers.openai import OpenAIProvider

provider = OpenAIProvider(
    base_url=f"{BASE_URL}/v1",
    api_key=AA_TOKEN,
)

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="qwen3-32b-tool",
    base_url=f"{BASE_URL}/v1",
    api_key=AA_TOKEN,
    use_responses_api=True,
)

curl $BASE_URL/health \
  -H "Authorization: Bearer $AA_TOKEN"

import httpx

response = httpx.get(
    f"{BASE_URL}/health",
    headers={"Authorization": f"Bearer {AA_TOKEN}"},
)
print(response.json())

curl -X POST $BASE_URL/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AA_TOKEN" \
  -d '{
    "model": "qwen3-32b-tool",
    "input": "What is the capital of Germany?"
  }'

response = client.responses.create(
    model="qwen3-32b-tool",
    input="What is the capital of Germany?",
)

print(response.id)           # "resp_abc123..."
print(response.status)       # "completed"
print(response.output_text)  # "The capital of Germany is Berlin."

agent = Agent(
    model=OpenAIResponsesModel("qwen3-32b-tool", provider=provider),
    system_prompt="You are a helpful assistant.",
)

result = await agent.run("What is the capital of Germany?")
print(result.output)  # "The capital of Germany is Berlin."

from typing import Annotated, TypedDict

from langchain_core.messages import BaseMessage, HumanMessage
from langgraph.graph import StateGraph
from langgraph.graph.message import add_messages

class State(TypedDict):
    messages: Annotated[list[BaseMessage], add_messages]
    previous_response_id: str | None

def chat(state: State) -> dict:
    kwargs = {}
    if pid := state.get("previous_response_id"):
        kwargs["previous_response_id"] = pid
    ai = llm.invoke(state["messages"], **kwargs)
    return {
        "messages": [ai],
        "previous_response_id": ai.response_metadata.get("id"),
    }

graph = (
    StateGraph(State)
    .add_node("chat", chat)
    .set_entry_point("chat")
    .set_finish_point("chat")
    .compile()
)

result = graph.invoke({
    "messages": [HumanMessage("What is the capital of Germany?")],
    "previous_response_id": None,
})
print(result["messages"][-1].text)  # "The capital of Germany is Berlin."
print(result["previous_response_id"])  # "resp_abc123...", pass to the next turn

curl -X POST $BASE_URL/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AA_TOKEN" \
  -d '{
    "model": "qwen3-32b-tool",
    "input": "Explain what a neural network is.",
    "instructions": "You are a helpful assistant that explains concepts in simple terms, using at most 2 sentences."
  }'

response = client.responses.create(
    model="qwen3-32b-tool",
    input="Explain what a neural network is.",
    instructions="You are a helpful assistant that explains concepts in simple terms, using at most 2 sentences.",
)

print(response.output_text)

agent = Agent(
    model=OpenAIResponsesModel("qwen3-32b-tool", provider=provider),
    system_prompt="You are a helpful assistant that explains concepts in simple terms, using at most 2 sentences.",
)

result = await agent.run("Explain what a neural network is.")
print(result.output)

from langchain_core.messages import HumanMessage, SystemMessage

result = graph.invoke({
    "messages": [
        SystemMessage(
            "You are a helpful assistant that explains concepts in simple terms, "
            "using at most 2 sentences."
        ),
        HumanMessage("Explain what a neural network is."),
    ],
    "previous_response_id": None,
})
print(result["messages"][-1].text)

curl -X POST $BASE_URL/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AA_TOKEN" \
  -d '{
    "model": "qwen3-32b-tool",
    "input": [
      {"type": "message", "role": "user", "content": "What is 5 + 3?"}
    ]
  }'

response = client.responses.create(
    model="qwen3-32b-tool",
    input=[
        {"type": "message", "role": "user", "content": "What is 5 + 3?"},
    ],
)

print(response.output_text)  # "8"

result = graph.invoke({
    "messages": [
        {"type": "message", "role": "user", "content": "What is 5 + 3?"},
    ],
    "previous_response_id": None,
})

print(result["messages"][-1].text)  # "8"