Integrations

Switch from a Hosted Gateway to AI24X

Already running on a hosted gateway? Keep your application as it is: one OpenAI-compatible key, smart routing and transparent pricing from $0.35/M.

Why developers are looking around

One endpoint for every model is now the default way teams build with LLMs. Once usage scales, three things decide the bill: pricing transparency, how many vendors you depend on, and how much of your spend goes to the gateway layer rather than to the models themselves.

AI24X is an independent, OpenAI-compatible gateway: one key for GPT-5, Claude, Gemini, DeepSeek, Kimi, Qwen and GLM, with smart routing that balances price, speed and stability — starting at $0.35/M (flash).

Migration in 4 steps (≈10 minutes)

1) Create an account and a key

# Register at https://www.ai24x.com/register # Create a key in the console → copy it once

2) Point your client to AI24X

Everything that speaks the OpenAI protocol works — just change the base URL and the key. For example with the OpenAI SDK:

from openai import OpenAI client = OpenAI( base_url="https://api.ai24x.com/v1", # your previous gateway endpoint api_key="sk-ai24x-..." # your AI24X key )

3) Map model names

# OpenAI-compatible model names are accepted and mapped automatically. # Short aliases also work: "flash" | "pro" | "vip-kimi" | "vip-mimo" ... # Unknown names fall back to flash (no hard failures).

4) Verify with one request

curl -sS -X POST "https://api.ai24x.com/v1/chat/completions" \ -H "Authorization: Bearer $AI24X_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"flash","messages":[{"role":"user","content":"Hello"}]}'

Expect HTTP 200 with an OpenAI-style response. 401 = wrong key · 402 = top up credits · 400 with model error = use a mapped model name.

What stays the same

  • Protocol: POST /v1/chat/completions, Bearer auth, streaming supported (SSE + [DONE]).
  • Your code: no SDK changes beyond base_url + key.
  • Model choice: keep calling the models you already use.

What changes for the better

  • Simpler pricing: flash from $0.35/M, pro tier for heavy reasoning, VIP picks for flagship Chinese models.
  • Smart routing: AI24X balances price, speed and stability per request — one bill, not per-provider surprises.
  • PayPal checkout: pay with PayPal, no credit card needed to start.

Get your key  ·  Hosted gateway vs AI24X comparison  ·  Pricing