AIVAX
Console

One endpoint. Shared AI behavior.

Applications keep one OpenAI-compatible request. Behind it, the model, knowledge, tools and policy can change as one shared system.

See how a request works

Two operating modes

Choose where the complexity should live.

Direct model

Keep a small call small.

Use an integrated model tag when the application owns the whole request and nothing needs to be shared yet.

Find an AI model
AI Gateway

Let shared behavior live behind the call.

Change instructions, retrieval, tools, input moderation or provider settings without redeploying every application.

Read the gateway guide

How a request works

Send the question. Keep the setup in one place.

Your application sends a message to a model or an AI Gateway. A gateway adds the shared behavior you configured, so the application does not have to rebuild it with every call.

Three applications. One shared configuration.
  • Web application
  • Mobile application
  • Backend service
One AI Gateway Instructions · Knowledge · Tools
Model · Output · Policy
Change it here.
Reuse it everywhere.

Each application keeps its own conversation while using the same configured behavior.

Illustrative clients. Configuration is shared; messages and responses remain separate.

A simple example

Turn an incident report into a next step.

An application reports a checkout timeout and asks for a severity, a summary and a next action. The response identifies a high-severity incident and suggests inspecting slow queries and database pool wait time.

See the API request and example response

Replace <gateway-slug> with your gateway identifier. Gateway slugs require a private API key; public keys must use the gateway full ID. The response and token counts below are illustrative, not a live run.

POST /v1/chat/completions 
curl https://inference.aivax.net/v1/chat/completions \
  -H "Authorization: Bearer $AIVAX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<gateway-slug>",
    "messages": [
      {
        "role": "user",
        "content": "Checkout timed out after the database pool reached capacity. Return severity, summary and next_action."
      }
    ],
    "response_schema": {
      "type": "object",
      "properties": {
        "severity": { "type": "string" },
        "summary": { "type": "string" },
        "next_action": { "type": "string" }
      },
      "required": ["severity", "summary", "next_action"]
    }
  }'

Example response content

{
  "severity": "high",
  "summary": "Checkout timeouts followed database pool saturation.",
  "next_action": "Inspect slow queries and database pool wait time."
}

Illustrative usage: 318 prompt tokens + 64 completion tokens = 382 total tokens.

After the answer

Understand what happened, not just what came back.

When conversation logging is enabled, the request and its operational context stay together. Use that record to investigate an unexpected answer, a cost increase or a failed call.

Open the Console

What you can inspect

  • The model and route

    See where the request originated and which model actually handled it.

  • The context behind the answer

    Follow the gateway, tools and linked collections that shaped the response.

  • Usage and errors

    Compare prompt and completion tokens, then inspect messages and errors when something goes wrong.

View the recorded fields

Origin and request ID connect the call to its record. The record includes the model name, gateway and API key context, linked resources, prompt/completion/total usage and token count, error message, messages, tools and metadata. Use idempotency_key and metadata to correlate records with your own workflow.

Connect another client

Let an MCP client call the gateway.

Publish one configured gateway as a callable tool without exposing the prompt, retrieval or provider setup behind it.

View MCP connection details
https://inference.aivax.net/v1/mcp/inference
Authorization
Bearer API key
X-Mcp-Model-Name
Published tool name
Read AI Gateway docs

Gateway and execution capabilities

02 / Reason

Shape how context becomes a reliable answer.

Gateway capabilities

03 / Act

Connect models and tools to systems outside the prompt.

Tools + providers / 08 capabilities

Change the system behind the endpoint, not every application.

Build in Console