OpenAI

Access every OpenAI model — GPT-4o, GPT-4.1, o3, o4-mini, DALL·E, Whisper, and embeddings — through TheRouter's unified API. Drop in your existing OpenAI SDK; just change one line to point base_url at TheRouter. Unified billing, automatic fallback, and team access controls included.

• ✓51+ models - the broadest production-ready portfolio in the industry
• ✓GPT-4o multimodal family with vision, audio, and real-time capabilities
• ✓o-series reasoning models for math, code, and complex problem solving
• ✓Responses API support for Cursor IDE and agent frameworks

Quickstart

from openai import OpenAI

# Access all OpenAI models through TheRouter — change base_url, keep everything else
client = OpenAI(
    base_url="https://api.therouter.ai/v1",
    api_key="YOUR_THEROUTER_KEY",
)

response = client.chat.completions.create(
    model="openai/gpt-4o",  # or openai/o4-mini, openai/gpt-4.1, openai/o3 ...
    messages=[{"role": "user", "content": "Summarize the key ideas in quantum computing"}],
    max_tokens=512,
)
print(response.choices[0].message.content)

# TypeScript / Node.js equivalent:
# import OpenAI from "openai";
# const client = new OpenAI({ baseURL: "https://api.therouter.ai/v1", apiKey: "YOUR_THEROUTER_KEY" });
# const res = await client.chat.completions.create({ model: "openai/gpt-4o", messages: [...] });

Use the global api.therouter.ai endpoint for new integrations; the legacy China accelerated endpoint is retired.