Integrations

Almost every tool that speaks OpenAI's API can use this one. You need three values:

Use one key per tool, so you can limit or delete one without touching the others. Setting labels below are taken from each tool's own docs; they can move between versions.

Coding agents

Claude Code

export ANTHROPIC_BASE_URL="https://pinerouter.ai"
export ANTHROPIC_AUTH_TOKEN="sk-..."
export ANTHROPIC_MODEL="gpt-5-2"

Full guide: Claude Code.

Codex CLI

# ~/.codex/config.toml
model = "gpt-5-5"
model_provider = "ai"

[model_providers.ai]
name = "AI"
base_url = "https://pinerouter.ai/v1"
env_key = "AI_API_KEY"
wire_api = "responses"

Full guide: Codex CLI.

Cursor

In Cursor Settings > Models, paste your key in the OpenAI API Key field, turn on the base URL override and set it to https://pinerouter.ai/v1, then add a model id such as gpt-5-2 by hand. Cursor uses your own key for chat models only; Tab completion stays on Cursor's models. Cursor sends these requests from its own servers.

Cline, Roo Code

API Provider: OpenAI Compatible, then:

Kilo Code

Add a custom provider: Provider API OpenAI Compatible, Base URL https://pinerouter.ai/v1, API key sk-..., then add model ids under Models (or pick from the fetched list).

opencode

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "ai": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "AI",
      "options": { "baseURL": "https://pinerouter.ai/v1", "apiKey": "{env:API_KEY}" },
      "models": { "gpt-5-2": { "name": "GPT 5.2" } }
    }
  }
}

Save it as opencode.json in your project.

Chat apps

Cherry Studio, CC Switch, DeepChat

Step-by-step guides: Cherry Studio, CC Switch, DeepChat.

LobeChat (self-hosted)

OPENAI_API_KEY=sk-...
OPENAI_PROXY_URL=https://pinerouter.ai/v1

Open WebUI

OPENAI_API_BASE_URL=https://pinerouter.ai/v1
OPENAI_API_KEY=sk-...

Or add the same URL and key as an OpenAI API connection in the admin settings under Connections.

Chatbox

Settings > Model Provider > Add, type OpenAI API compatible:

Add at least one model id, save and click Check.

Frameworks

Vercel AI SDK

import { createOpenAICompatible } from "@ai-sdk/openai-compatible";
import { generateText } from "ai";

const ai = createOpenAICompatible({ name: "ai", baseURL: "https://pinerouter.ai/v1", apiKey: process.env.API_KEY });
const { text } = await generateText({ model: ai("gpt-5-2"), prompt: "Hello" });

LangChain

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(model="gpt-5-2", base_url="https://pinerouter.ai/v1", api_key="sk-...")
print(llm.invoke("Hello").content)

ChatOpenAI reads only fields in OpenAI's standard format, so model-specific extras in answers are dropped.

LlamaIndex

# pip install llama-index-llms-openai-like
from llama_index.llms.openai_like import OpenAILike

llm = OpenAILike(model="gpt-5-2", api_base="https://pinerouter.ai/v1", api_key="sk-...",
                 is_chat_model=True, context_window=128000)

Set context_window to the model's real context length.

LiteLLM

import litellm

r = litellm.completion(model="openai/gpt-5-2", api_base="https://pinerouter.ai/v1", api_key="sk-...",
                       messages=[{"role": "user", "content": "Hello"}])

The openai/ prefix tells LiteLLM to use the OpenAI format; the part after it is our model id.

Automation

n8n

Create an OpenAI credential: API Key sk-..., Base URL https://pinerouter.ai/v1. Use it in the OpenAI and Chat OpenAI nodes and type the model id.

Dify

Install the OpenAI-API-compatible model provider, then add a model: Model Name gpt-5-2, API Key sk-..., API endpoint URL https://pinerouter.ai/v1.

Flowise

Use the ChatOpenAI node with an OpenAI credential holding your key, set Model Name to an id such as gpt-5-2, and under Additional Parameters set Base Path to https://pinerouter.ai/v1.

MCP clients

Claude.ai, ChatGPT, Claude Code, Codex, Cursor and other MCP clients can call the media and text models as tools through https://pinerouter.ai/mcp, with sign-in or your key. See MCP.

Media models in these tools

These tools speak the text formats. For images, video, music and speech with every setting, call Jobs from your code, or use the MCP tools above.