What do we do?

We function as an inference gateway for various models from various providers. We centralize useful functions for creating specialist agents.

You can use your own OpenAI compatible model through your own API-Key too.

  • Tuning
    Your gateways already deliver your ready-made model with specific instructions, hyper-parameterization, prefilling and functions.
  • RAG-optimized
    Create collections of thousands of documents indexed in your assistant, optimizing answers with large amounts of specific knowledge.
  • Connected functions
    Provide functions directly from your API so your assistant can communicate with existing services and APIs.
  • Agent tools
    Use our ready functions such as internet search, access to external links, code execution and memory for assistants without touching a single line of code.
Integration in minutes

Connect your gateway to a functional and elegant chat client.

Your client is connected to your users with a few lines of code, where it is possible to manage context and limits.

Transparent and Simple Pricing
You only pay for what you use. No hidden fees.
Pay-per-use

We charge the same token prices as the providers, without any extra fees. Your cost is directly tied to your usage.

Directly from the source.
Service Fee

To keep the platform running, we add a small fee that covers taxes and operational costs when you add credits to your account.

For platform maintenance.
Connect Your AI to the World

Use our functions to specify what you want in your input JSON, and let the assistant handle the rest. AIVAX enables your model to search the internet, execute code, and access links to deliver exactly what you need. 

// User request
> {
    "instructions": "I want the latest SELIC and IPCA rates for today",
    "responseFormat": { "ipcaToday": 0, "selicToday": 0 }
}

// AIVAX response
< { "ipcaToday": 5.2, "selicToday": 14.0 }
Be part of it

Be part of AIVAX. The product is in beta, but we are accepting invitations to join.

Join now