Quick Start

Get up and running in under 5 minutes. Aqta is a drop-in replacement for the OpenAI base URL — one line of config, full governance.

1. Get your API key

Sign in at app.aqta.ai and go to Settings → API Keys → New key. Copy the key — it starts with sk-aqta-.

2. Point your client at Aqta

Change one line in your existing code:

Python (openai SDK)

pythonCopy
from openai import OpenAI

client = OpenAI(
    base_url="https://api.aqta.ai/v1",
    api_key="sk-aqta-your-key-here",
)

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Summarise this contract in 3 bullet points."}],
)

print(response.choices[0].message.content)
# aqta metadata: trace_id, cost_eur, policy outcome
print(response.aqta)

TypeScript / Node.js

typescriptCopy
import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'https://api.aqta.ai/v1',
  apiKey: 'sk-aqta-your-key-here',
});

const response = await client.chat.completions.create({
  model: 'gpt-4o',
  messages: [{ role: 'user', content: 'Summarise this contract in 3 bullet points.' }],
});

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

cURL

bashCopy
curl https://api.aqta.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-aqta-your-key-here" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

3. Add your provider key

Aqta proxies to OpenAI, Anthropic, Google, Mistral, Cohere, Groq, and Perplexity. Add your provider credentials once in the dashboard — Aqta stores them encrypted (AES-256) and injects them per-request. Your application never handles provider keys directly.

Go to Settings → Integrations → Provider vault and add the keys for the providers you use.

4. Verify in the dashboard

Every request appears in Activity within seconds, showing:

• Trace ID and timestamp
• Model and provider used
• Token counts and cost in EUR
• Policy outcome: passed, suppressed, or blocked

What happens under the hood

Every request goes through the Aqta policy engine before reaching your provider:

1. Loop detection — detects runaway agent loops, breaks the circuit before costs compound
2. Budget enforcement — blocks requests that would exceed your configured spend limits
3. Policy check — validates model, content type, and use-case against your org rules
4. Cryptographic attestation — signs a receipt for every request; tamper-evident, exportable for audit

The total overhead is under 3 ms at P99.

Next steps

• Authentication — manage API keys, rotation, and security
• Endpoints — full API reference
• Rate Limits — limits by tier and retry guidance
• Compliance — GDPR, EU AI Act, HIPAA coverage

Questions? hello@aqta.ai