Quickstart

Make your first request against the ResetData inference API in a few minutes. The API is OpenAI-compatible, so existing OpenAI SDKs work by pointing the base URL at ResetData.

1. Get an API key. Create one from API keys in the ResetData app, then export it. Keys start with sk-rd- and are shown in full only once.

   export RESETDATA_API_KEY="sk-rd-..."

2. Send a chat completion. The endpoint is https://api.resetdata.ai/api/v1/chat/completions.

   cURL

   curl https://api.resetdata.ai/api/v1/chat/completions \
     -H "Authorization: Bearer $RESETDATA_API_KEY" \
     -H "Content-Type: application/json" \
     -d '{
       "model": "zai/glm-5.2",
       "messages": [
         { "role": "user", "content": "Hello from ResetData!" }
       ]
     }'

   Python

   import os
   
   from openai import OpenAI
   
   client = OpenAI(
       base_url="https://api.resetdata.ai/api/v1",
       api_key=os.environ["RESETDATA_API_KEY"],
   )
   
   resp = client.chat.completions.create(
       model="zai/glm-5.2",
       messages=[{"role": "user", "content": "Hello from ResetData!"}],
   )
   print(resp.choices[0].message.content)

   JavaScript

   import OpenAI from "openai";
   
   const client = new OpenAI({
     baseURL: "https://api.resetdata.ai/api/v1",
     apiKey: process.env.RESETDATA_API_KEY,
   });
   
   const resp = await client.chat.completions.create({
     model: "zai/glm-5.2",
     messages: [{ role: "user", content: "Hello from ResetData!" }],
   });
   console.log(resp.choices[0].message.content);

3. Read the response. You’ll get an OpenAI-style chat.completion object with the assistant’s reply in choices[0].message.content.

Models are not interchangeable

Swap zai/glm-5.2 for any model in the catalog — but check that model’s page first. Every model declares its own set of supported parameters, and anything outside that set is silently dropped before the request reaches the model. Your call will succeed and simply ignore the parameter.

Sampling options, tool-calling support, reasoning effort and context length all vary between models. See Model parameters before you tune a request.

Next

Authentication for how keys are scoped · Choosing a model for where to start · Errors for handling 429 and 503 · Rate limits and queueing before you go to production.