Quickstart

Infer is a unified OpenAI-compatible endpoint for top AI models. This page walks you through the four things you need to start using it: an API key, the base URL, a connection test, and the model IDs you can call.

Get an API key

Open the API Keys dashboard.
Sign in or create an Infer account.
Click Create API key, name it, and copy the generated key (format: sk-infer-...).
Confirm the team has available balance on the Billing page so the key can make paid calls.

Warning:

Keys are shown only once. Store them in a password manager or server-side secret store. Never commit them to source control or paste them into client-side code.

Base URL

Every Infer request goes to the same OpenAI-compatible endpoint. Use this value for the base_url / baseURL field in any SDK or the target URL for direct HTTP calls:

Base URL

https://api-agenthub-pre.riema.xyz/v1

The URL above is resolved for the environment you are signed in to, so copy-paste works without further editing.

Verify the connection

Run one of the snippets below with your key in place of your_api_key. A successful call returns a chat completion JSON payload, which confirms the key, the base URL, and your network are all working:

curl https://api-agenthub-pre.riema.xyz/v1/chat/completions \
  -H "Authorization: Bearer your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "your_model",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

A successful response looks like this — the choices[0].message.content is the model’s reply, and the usage block shows the tokens billed against your balance:

200 OK

{
"id": "chatcmpl-9f2abc123",
"object": "chat.completion",
"created": 1716240000,
"model": "gpt-5.4",
"choices": [
  {
    "index": 0,
    "message": {
      "role": "assistant",
      "content": "Hello! How can I help you today?"
    },
    "finish_reason": "stop"
  }
],
"usage": {
  "prompt_tokens": 9,
  "completion_tokens": 11,
  "total_tokens": 20
}
}

Tip:

A 401 means the key is wrong or missing. A 404 on /chat/completions almost always means the base URL is missing the /v1 suffix; copy it from the block above.

Supported models

These are the model IDs you can use to test the connection — drop any one of them into the model field of the verification request above to check whether that specific model is currently callable on your account. The full catalog with live availability lives on the Models page.

Vendor	Model ID
OpenAI	`gpt-5.4gpt-5.4-mini`
Anthropic	`claude-opus-4-6claude-sonnet-4-6`
Google	`gemini-3.1-pro-previewgemini-3.1-flash-lite-preview`
Alibaba	`qwen3.5-plusqwen3.5-flashqwen3.6-plusqwen3-coder-plus`
DeepSeek	`deepseek-v3.2`
MiniMax	`MiniMax-M2.5`
Moonshot	`kimi-k2.5`
Zhipu	`glm-5`

Supported tools

Infer works with any client that accepts a custom OpenAI-compatible base URL. Pick the integration path that matches how you use AI, and follow its setup guide:

Quickstart

Create an API key, point your client at Infer, and make the first request.

CLI

Set up terminal agents such as Codex, Claude Code, and OpenCode with environment variables and config files.

VS Code extension

Use VS Code-based AI extensions with Infer as the model gateway.

API

Call the OpenAI-compatible API directly from Python, Node.js, curl, or any HTTP client.

CLI agents: terminal agents including Codex, Claude Code, and OpenCode.
VS Code extensions: the first-party Codex plugin and other extensions that read a custom OpenAI endpoint.
Direct API: any application, SDK, or script that speaks the OpenAI /chat/completions protocol. See the API reference for full details.

What’s next

Base URL for this environment: https://api-agenthub-pre.riema.xyz/v1.
Manage keys and view usage from the API Keys dashboard.
Browse the full Models catalog with live availability.
Read the API reference for authentication, error codes, and privacy.