Getting Started

Quick Start

VectorAxis exposes an OpenAI-compatible API. You only need to change your base URL and use a VectorAxis virtual key. The rest of your code stays identical. Four steps and you’re live.

Create a virtual key

In the dashboard, go to Virtual Keys → New Key. Choose a provider (e.g. OpenAI), paste your provider API key, and optionally set a credit budget or rate limits. Copy the returned slug (vk-…). That’s what your app will use instead of the real key. No provider account of your own? Create a PLATFORM key and draw on prepaid credits instead. Full detail on the Virtual Keys page.

Point your SDK at VectorAxis

Replace your provider base URL with the gateway URL and set api_key to your virtual key slug. The SDK sends it as Authorization: Bearer and VectorAxis routes to the right provider.

python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.vectoraxis.ai/v1",
    api_key="vk-1a2b3c4d5e6f7a8b",   # your VectorAxis virtual key slug
)

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
)

Or with plain HTTP:

bash

curl https://api.vectoraxis.ai/v1/chat/completions \
  -H "Authorization: Bearer vk-1a2b3c4d5e6f7a8b" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4o","messages":[{"role":"user","content":"Hello!"}]}'

Add gateway features via headers

Every VectorAxis capability is activated by a request header, with no SDK changes required. Turn on caching, retries, and a fallback provider in one call:

python

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[...],
    extra_headers={
        "x-cache-mode": "semantic",          # serve similar queries from cache
        "x-retry-count": "3",                # retry transient failures
        "x-fallback-1-provider": "ANTHROPIC",  # fail over to Anthropic
        "x-fallback-1-api-key": "sk-ant-...",
        "x-guardrail": "pii-redact",         # redact PII on the way out
    },
)

See every header on the Request Headers reference.

View logs and cost in the dashboard

Every request is logged automatically. Open Logs to inspect latency, tokens, cost, cache status, and the full request/response bodies; open Analytics for cost and latency trends across all traffic. (Attribution to an org/workspace is automatic when you call with a virtual key.)

Where to next? Spread traffic with Routing Configs, cut cost with Caching, enforce safety with Guardrails, or wire up your IDE via Coding Agents.

VectorAxis is built by GK Labs, a member of NVIDIA Inception.