Your models. One route.
Build with Viki.

Route your AI requests through one simple API. Choose from 50
text, image, and video models. One key. One shared balance.

Example route: Claude. You choose the model.
Many models.
One API.
OpenAI logoOpenAIAnthropic logoAnthropicGoogle logoGoogleFLUX logoFLUXxAI logoxAIKling logoKling
50 models

More models should
mean more possibilities.
Not more integrations.

A new provider. Another integration.

Different keys, request formats, and accounts. Each provider adds another moving part to your application.

Your requests, spread across accounts.

Separate balances and billing pages make it harder to follow what your AI features actually cost.

You choose the model.
Viki connects the dots.

Connect to multiple providers through one API, with the model choice in your hands.

One key. Every model.

Connect your application with one Viki API key. Keep the same integration as you try different models.

Explore the API

Text, image, and video.

Write, create, and generate with 50 models in one library. Choose the model that fits what you want to build.

Browse 50 models

One shared balance.

Use one prepaid balance across your chosen models. Track each request and its cost, including Viki’s 20% platform fee.

Understand pricing

Your request.
Your chosen model.

Pick a model from the library and include its ID in your request. Viki routes it to your chosen model, using your Viki key and balance.

Explore models
An example route
You choose the modelmodel: "openai/gpt-6.1-sol"
Your application
Viki API
Your selected model
OpenAI logoGPT-6.1 SolYour selected model · OpenAI

A familiar API.
A simpler route.

Keep your OpenAI SDK for text generation. Update the base URL and API key, choose your model, and make your first request.

Read the docs
hello-viki.ts
import OpenAI from "openai";

const viki = new OpenAI({
  apiKey: process.env.VIKI_API_KEY,
  baseURL: "https://vikiapi.com/v1",
});

const response = await viki.chat.completions.create({
  model: "openai/gpt-6.1-sol",
  messages: [
    { role: "user", content: "Let's build something." }
  ],
});
OpenAI-compatible text API

Many providers.
One shared balance.

Add credits once and use them across your chosen models. Every request includes the provider cost plus Viki's 20% platform fee.

View pricing
One request. A clear cost.
Provider model cost$1.00
Viki platform fee · 20%$0.20
From your balance$1.20

Example calculation. Actual charges depend on your chosen model and request.

Choose your model.
Build your next idea.

What is Viki?

Viki is one API for text, image, and video models. Choose a model in your request, and Viki routes it to your selected model. One API key and prepaid balance connect your application to multiple providers.

Who chooses the model?

You do. Include the model ID in your request, and Viki sends it to that model. You can choose a different model for each request. Viki does not automatically select a model for you.

Which models can I use?

Choose from 50 models: 12 text models, 19 image models, and 19 video models. Browse the model library for the current providers, capabilities, and estimated rates. Availability depends on the upstream provider.

How does pricing work?

Buy prepaid credits and pay for the requests you make. Viki adds a 20% platform fee to the provider's actual model cost. There is no monthly subscription. Your dashboard records usage and charges.

Can I use my existing OpenAI SDK?

Yes. For text generation, set your SDK's base URL to your Viki deployment followed by /v1, and use a Viki API key. Image and video generation use their own endpoints. Find complete examples in the documentation.

How do I get started?

Create an account, add credits, and generate an API key. Choose a model from the library and connect your application using the documentation. You can also try models in the playground. Generation and credit purchases will be available once service setup is complete.