API
The Ox Alpha API has been retired
Ox Alpha's free, OpenAI-compatible endpoint served GLM-5.3 Flash until September 30, 2026. It no longer answers, but the model is only a base URL away: the same OpenAI SDK code works against any of the APIs below.
If your code still calls Ox Alpha
- Requests to
https://oxalpha.run/api/v1/chat/completionsreturn410 Gonewith the error codeapi_retired. The official OpenAI SDKs do not retry a 410, so nothing loops. - Ox Alpha keys no longer work anywhere. There is nothing to revoke or delete on your side.
GET /api/statusnow answers"state": "retired"for good.- To switch, change three values: the base URL, the key, and the model id.
Where to call it now
Three OpenAI-compatible APIs that serve GLM-5.3 Flash
Z.ai, the maker
Z.ai serves its own model through an OpenAI-compatible API at https://api.z.ai/api/paas/v4/. Create a key in your Z.ai account; the model name and its options are in Z.ai's GLM-5.3 Flash guide, and the setup in its API quick-start.
OpenRouter
One key, many hosts: set the base URL to https://openrouter.ai/api/v1 and the model to z-ai/glm-5.3-flash. Several hosts charge less than Z.ai; the live price of every host is on our GLM-5.3 Flash page.
GenMagic
GenMagic, a multi-model studio the team behind Ox Alpha also runs, has an OpenAI-compatible API with streaming at https://genmagic.co/api/v1, where you pick GLM-5.3 Flash from its model list. Calls draw on the same credit balance as the studio, at the same price.
Would rather run it yourself? GLM-5.3 Flash is open weight under the MIT license, with the weights on Hugging Face (zai-org), and serves with vLLM or SGLang. Every way to run it
The change in code
Three values, nothing else
Shown with OpenRouter because its model id is fixed and public. For Z.ai or GenMagic, swap in their base URL, your key there, and their name for the model.
| Setting | Ox Alpha (retired) | OpenRouter |
|---|---|---|
| base URL | https://oxalpha.run/api/v1 | https://openrouter.ai/api/v1 |
| model | ox-alpha | z-ai/glm-5.3-flash |
| key | your Ox Alpha key | your OpenRouter key |
from openai import OpenAI
client = OpenAI(
base_url="https://openrouter.ai/api/v1", # was https://oxalpha.run/api/v1
api_key="YOUR_OPENROUTER_KEY", # was your Ox Alpha key
)
res = client.chat.completions.create(
model="z-ai/glm-5.3-flash", # was "ox-alpha"
messages=[{"role": "user", "content": "Explain your 1M context in one line."}],
)
print(res.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1", // was https://oxalpha.run/api/v1
apiKey: process.env.OPENROUTER_API_KEY, // was your Ox Alpha key
});
const res = await client.chat.completions.create({
model: "z-ai/glm-5.3-flash", // was "ox-alpha"
messages: [{ role: "user", content: "Explain your 1M context in one line." }],
});
console.log(res.choices[0].message.content);Skip the wiring: build on Founden
An API is for calling the model from your own code. If you would rather go straight from idea to a working product, company, or 3D world without standing up the infrastructure, Founden builds it for you.