Documentation

Get your API key

Create an account on our platform and generate your first API key. A single key that works with all models and compatibility formats.

First API call

Use the OpenAI-compatible endpoint to make your first request. Replace $APIARIO_KEY with your key.

curl https://api.apiario.dev/v1/chat/completions \
  -H "Authorization: Bearer $APIARIO_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "maritaca/sabia-4",
    "messages": [
      {"role": "user", "content": "Olá! Quem é você?"}
    ]
  }'

from openai import OpenAI

client = OpenAI(
    base_url="https://api.apiario.dev/v1",
    api_key="$APIARIO_KEY"
)

response = client.chat.completions.create(
    model="maritaca/sabia-4",
    messages=[
        {"role": "user", "content": "Olá! Quem é você?"}
    ]
)

print(response.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.apiario.dev/v1",
  apiKey: process.env.APIARIO_KEY,
});

const response = await client.chat.completions.create({
  model: "maritaca/sabia-4",
  messages: [
    { role: "user", content: "Olá! Quem é você?" }
  ],
});

console.log(response.choices[0].message.content);

Models & Pricing

Browse all available models via GET /v1/models. Prices are displayed in BRL per 1 million tokens, for both input and output.

curl https://api.apiario.dev/v1/models \
  -H "Authorization: Bearer $APIARIO_KEY"

Tokens & Usage

Each request consumes credits proportionally to the number of tokens processed. Track your usage in real time on the dashboard. 1 credit = R$ 0.001.

• 1 crédito = R$ 0,001
• Each request debits credits from your balance in real time
• Track on the dashboard: daily usage, by model, period totals
• Export reports in CSV

Rate Limit & Isolation

Each upstream provider (OpenAI, Anthropic, Maritaca, etc.) sets its own rate limits (RPM, TPM). Apiário Dev transparently relays these limits. Each API key has its own isolation — check response headers to monitor your consumption.

Error Codes

The API uses standard HTTP status codes. Validation errors return 422, insufficient credits return 402, and upstream provider errors are forwarded in the appropriate format (OpenAI, Anthropic, or Ollama).

Next steps

Explore the dashboard to manage your keys, track usage, and top up credits.

Frequently Asked Questions

Common questions about the Apiário Dev API.