Quickstart

NoviaHub speaks the OpenAI, Anthropic and Gemini protocols. Code that already uses one of those official SDKs can call models on NoviaHub by changing only the address and the API key.

1. Create an API key

   Sign in at noviahub.com (see Signing in and your account), open API Keys in the console, click Create API Key, enter a name and click Save changes. Then click the masked key in the list and copy the full key (it starts with sk-). API keys explains every option.

   Usage is charged to your account balance, and requests fail once the balance runs out. You can top up in Wallet; see Wallet and top-ups.

2. Put the key in an environment variable

   Keeping the key in an environment variable lets code and tools read it without writing the key into your code. The commands below only apply to the current terminal window; to keep the variable, add the line to the end of ~/.zshrc or ~/.bashrc.

   macOS / Linux

   export NOVIAHUB_API_KEY="sk-..."

   Windows PowerShell

   $env:NOVIAHUB_API_KEY = "sk-..."

   Variable name

   These docs always use the name NOVIAHUB_API_KEY. The samples you copy from a model’s API access section on the NoviaHub website use NEW_API_KEY instead. Only the name differs; pick one and use it consistently.

   Caution

   Anyone holding the key can spend your balance. Keep it out of frontend code, repositories and chats. If it leaks, delete it on the API Keys page and create a new one.

3. Send your first request

   This calls deepseek-v4-flash through the OpenAI-compatible Chat Completions API. To use another model, change model; model IDs are listed under Models & Pricing.

   cURL

   curl https://noviahub.com/v1/chat/completions \
     -H "Authorization: Bearer $NOVIAHUB_API_KEY" \
     -H "Content-Type: application/json" \
     -d '{
       "model": "deepseek-v4-flash",
       "messages": [{"role": "user", "content": "Introduce yourself in one sentence."}]
     }'

   Python

   import os
   
   from openai import OpenAI
   
   client = OpenAI(
       base_url="https://noviahub.com/v1",
       api_key=os.environ["NOVIAHUB_API_KEY"],
   )
   
   completion = client.chat.completions.create(
       model="deepseek-v4-flash",
       messages=[{"role": "user", "content": "Introduce yourself in one sentence."}],
   )
   
   print(completion.choices[0].message.content)

   TypeScript

   import OpenAI from 'openai'
   
   const client = new OpenAI({
     baseURL: 'https://noviahub.com/v1',
     apiKey: process.env.NOVIAHUB_API_KEY,
   })
   
   const completion = await client.chat.completions.create({
     model: 'deepseek-v4-flash',
     messages: [{ role: 'user', content: 'Introduce yourself in one sentence.' }],
   })
   
   console.log(completion.choices[0].message.content)

   The answer is in choices[0].message.content. Each call’s usage and cost appear under Usage Logs in the console.

Addresses

All three protocols use the same API key. They differ in the address and in the header that carries the key:

Protocol	SDK base URL	Auth header
OpenAI-compatible	https://noviahub.com/v1	Authorization: Bearer <key>
Anthropic Messages	https://noviahub.com	x-api-key: <key>; Authorization: Bearer <key> also works
Gemini	https://noviahub.com	x-goog-api-key: <key>

The Anthropic and Google Gen AI SDKs append /v1/messages and /v1beta/models/... themselves, so their base URL is the bare domain, without /v1.

To see which protocols a model supports, open it under Models & Pricing. Its API access section also has ready-made code samples and coding agent configs for that model.

Next steps

Claude CodeRoute Claude Code through NoviaHub.

CodexAdd NoviaHub as a model provider in Codex.

API referenceEndpoints, parameters and responses for each protocol.

BillingCharging rules, balance and usage records.