Use Redline with LangChain
LangChain's ChatOpenAI takes a custom base_url. One flag matters: turn the Responses API off, or LangChain may call an endpoint Redline does not serve.
Python
# pip install -U langchain-openai
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(
model="deepseek/deepseek-chat-v3.1",
api_key="rl-...",
base_url="https://redline-gateway.pages.dev/v1",
use_responses_api=False,
)
print(llm.invoke("Hello").content)
Keep use_responses_api False With a custom base URL LangChain can switch to the Responses API, which returns 404 on Redline. Set use_responses_api=False. If streaming usage misbehaves, it is off by default on a custom base URL, which is fine.
JavaScript
// npm install @langchain/openai @langchain/core
import { ChatOpenAI } from "@langchain/openai";
const llm = new ChatOpenAI({
model: "deepseek/deepseek-chat-v3.1",
apiKey: process.env.REDLINE_API_KEY,
configuration: { baseURL: "https://redline-gateway.pages.dev/v1" },
});
console.log((await llm.invoke("Hello")).content);