LLM API gateway for developers

One API.All models.Best price.

Access GPT, Claude, and Gemini from Russia without VPN. One OpenAI-compatible endpoint, ruble balance, and documents for companies.

Get API key See pricing

BASE_URL=https://app.promptra.ru

curl "$BASE_URL/v1/chat/completions" \
  -H "Authorization: Bearer sk_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.4",
    "stream": true,
    "messages": [
      {"role": "user", "content": "Reply briefly"}
    ]
  }'

Compatible LLM providers

Catalog

Flagship models without manually choosing providers.

Promptra keeps public model ids stable while routing, availability, and upstream accounts are handled inside the platform.

providermodelcontextlatencyinputoutput

Anthropicclaude-opus-4.7400K1.2 с$5.00$25.00

OpenAIgpt-5.4272K0.9 с$1.25$10.00

Googlegemini-3-pro1M1.4 с$0.35$2.10

DeepSeekdeepseek-v3.2128K0.7 с$0.14$0.28

Pricing

Token prices match providers. Fee is charged only when topping up.

5%service fee when topping up

100 RUBminimum deposit for individuals

5,000 RUBstarting amount for company contracts

Comparison

Where we are different. And where we honestly lose.

One `/v1` endpointyesoften

Ruble paymentsyesnot always

Legal documentsyesrarely

Flagship model focusyesno

Hundreds of modelsnoyes

FAQ

Questions people ask first.

Is it compatible with the OpenAI SDK?

Yes. Clients, SDKs, and tools that support an OpenAI-compatible endpoint can point to Promptra `/v1`.

How does payment work?

The balance is topped up in rubles. For legal entities we support contracts, invoices, and closing documents.

Do I need a VPN?

No. The API is reachable from Russia and does not require foreign cards or a personal VPN.

Why fewer models than aggregators?

We focus on flagship models and stable routing. The catalog expands only after quality and availability checks.

Ready to try

One endpoint. All models. Ruble balance.

Create an account, get an API key, and connect any OpenAI-compatible client to Promptra.

Get API key