One line. Thirty models. Zero rewrite.

Why switch from OpenAI?

OpenAI is a great platform. But a single-provider dependency means your pricing, your features, and your uptime follow someone else's roadmap. Teams switch to Plugsky to get: multiple model families on one key, in-region hosting, automatic fallback when a model is overloaded, and pricing that does not require enterprise negotiation.

The one-line switch

Plugsky is OpenAI-compatible. You do not rewrite your code:

# Python — OpenAI SDK
from openai import OpenAI
client = OpenAI(
    base_url="https://api.plugsky.com/v1",
    api_key="sk-live-...",   # from your Plugsky dashboard
)
resp = client.chat.completions.create(
    model="plugsky-pro",
    messages=[{"role": "user", "content": "Hello!"}],
)

Same for the Node SDK, curl, LangChain, and every framework that talks OpenAI protocol.

Switch credit

Teams migrating from a paid AI provider can apply for a switch credit matching up to one month of their current bill (redacted invoice required). This covers the overlap period while you test and migrate.

Migration steps

1. Start the $5 trial and generate a key.
2. Change base_url in one environment.
3. Run your test suite against the same prompts.
4. Compare quality, latency, and cost.
5. Move remaining environments, then apply for the switch credit.

FAQ

Is every OpenAI feature supported?

Chat completions, streaming, embeddings, vision, and function calling are supported. See the compatibility report.

What about existing OpenAI model names in my code?

Change the model names to the Plugsky equivalents (e.g. gpt-4o → plugsky-pro). Model mapping is documented in the API reference.

How long does migration take?

Most teams finish in an afternoon for one environment, and a week for a full migration.