IranVia

AI Gateway

One API key for Claude, GPT, Gemini, DeepSeek, Llama and more, billed from your Novin Cloud wallet — no international card, no proxy.

Usage GuideOpen AI Console
AI Gateway

Using large language models from Iran has three recurring blockers: you need an international credit card, providers block access, and every provider means another account and key. Novin Cloud AI removes all three with one gateway.

What is Novin Cloud AI?

An AI gateway that puts dozens of models from different providers behind one standard API. Instead of opening accounts with OpenAI, Anthropic and Google and paying in foreign currency, one Novin Cloud account, one API key and your rial wallet give you every model. Available families include Claude (Opus, Sonnet, Haiku), GPT and the o-series reasoning models, Gemini with very large context windows, Llama, DeepSeek, Mistral and Codestral, and Qwen Coder. The live list with prices, context window and capabilities (vision, function calling, streaming) is in the console and at the standard /v1/models endpoint.

How it works

1. Your app sends the request with an Authorization: Bearer header to https://iapi.novin.cloud/v1. 2. The gateway authenticates the key and enforces its limits (allowed models, RPM, TPM, budget cap, expiry). 3. The request is routed to the right provider based on the model ID, e.g. anthropic/claude-sonnet-4-5 or openai/gpt-4o. 4. The response, streamed or not, comes back in the OpenAI format, and its usage block reports exact input and output tokens. 5. The cost of that request (input tokens × input price + output tokens × output price) is deducted from your wallet and recorded on the console usage page.

Benefits over connecting directly

One account instead of one per provider. One API key instead of several keys and invoices. Rial wallet billing instead of an international card. Direct access from Iran with no middleman or proxy. One standard API instead of per-provider code; switching models is a one-string change. Cost guard: if a key leaks or code loops forever, the loss stays capped at that key's budget.

Two ways to use it: API and chat

Through the API, any app, script or tool that speaks OpenAI connects by changing two values — from cURL and the official Python and Node.js SDKs to streaming, JSON output and function calling. Without code, use the console chat to test and compare models quickly, or chat.novin.cloud with the same account and server-side history on any device.

Summary

If you use the OpenAI SDK today, moving to Novin Cloud is a two-line change. Top up your wallet, create a key with a budget cap and send your first request in under five minutes. The full guide is in the Novin Cloud AI docs.

Use cases

One key for every model

Claude, GPT, Gemini, Llama, DeepSeek, Mistral and Qwen with one API key and one endpoint: iapi.novin.cloud

Drop-in OpenAI compatibility

Change only base_url and api_key; LangChain, LlamaIndex, Cursor and Continue work without code changes.

Pay per use in rials

No subscription or minimum; cost is charged per input and output token from your wallet.

Full cost and access control

Budget cap, reset period, RPM/TPM limits, allowed models and expiry on every key.

Chatbots and intelligent customer support

Coding assistants in Cursor, Continue and VS Code

Summarizing, translating and analyzing Persian and English documents

Agents with function calling and JSON output

Start in under five minutes

From key to first model response; all you need is a Novin Cloud account and wallet balance.

  1. 1

    Create an API key

    In the console go to AI → API Keys and click “Create new key”. The key is shown only once; store it in an environment variable:

    export NOVIN_API_KEY="sk-xxxxxxxxxxxxxxxxxxxxxxxx"
  2. 2

    First request with cURL

    Model IDs follow the provider/model-name pattern:

    curl https://iapi.novin.cloud/v1/chat/completions \
  -H "Authorization: Bearer $NOVIN_API_KEY" \
  -H "Content-Type: application/json" \
  -d {"model": "openai/gpt-4o", "messages": [{"role": "user", "content": "Hello!"}]}
  3. 3

    Official OpenAI SDK in Python

    Your existing code works by changing api_key and base_url:

    from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ["NOVIN_API_KEY"],
    base_url="https://iapi.novin.cloud/v1",
)
resp = client.chat.completions.create(
    model="anthropic/claude-sonnet-4-5",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)
  4. 4

    List available models

    Get exact model IDs from the standard endpoint:

    curl https://iapi.novin.cloud/v1/models \
  -H "Authorization: Bearer $NOVIN_API_KEY"

Full API reference, streaming, function calling and error codes are in the docs: docs.novin.cloud/ai/about

Service links

Related console sections and official documentation.

Frequently Asked Questions

Send your first request now

Create an account, top up your wallet and get an API key. The full guide is at docs.novin.cloud/ai/about.

Usage GuideOpen AI Console