I used to have 4 API keys, 4 billing dashboards, and 4 SDKs. Now I have one string.

I used to have 4 API keys, 4 billing dashboards, and 4 SDKs. Now I have one string.

import openai
client = openai.OpenAI(
api_key="mb-xxx",
base_url="https://aibridge-api.com/v1"
)

That's the setup. Everything else is just changing the model name:

DeepSeek — fast, cheap ($0.27/M)
client.chat.completions.create(model="deepseek-chat", messages=msgs)

Qwen3 — 128K context, multilingual
client.chat.completions.create(model="qwen3-235b-a22b", messages=msgs)

GLM-4 Plus — complex reasoning
client.chat.completions.create(model="glm-4-plus", messages=msgs)

Kimi K3 — 1M context, always reasoning
client.chat.completions.create(model="kimi-k3", messages=msgs)

Same library. Same response format. Same key. Different model string. That's the entire migration from "I call one provider" to "I call whichever model fits the task."

What ships with it:

GitHub OAuth login. Click GitHub → authorize → dashboard. Five seconds. Registration pages that used to ask for email, password, and a verification code now have one button.

A free playground at aibridge-api.com/playground.html. No signup. 10 requests/day. Pick a model, type a prompt, see the response. If you like it, the "Register" button is right there.

A dashboard that actually tells you where you stand — blue progress bar below 50%, orange at 80%, red with a flashing warning at 90%, plus an upgrade prompt that stops being polite and starts being visible exactly when you need it.

The stack:

15 models — DeepSeek (5), Qwen (3), GLM (3), Moonshot Kimi (4). One OpenAI-compatible endpoint. Streaming, function calling, JSON mode. Free tier: 500K tokens/month. No credit card.

→ aibridge-api.com
→ aibridge-api.com/playground.html (try all 15 models, no signup)