Skip to content
Apertis
Get started

OpenAI-compatible. Provider-independent.

One AI API gateway for 500+ models.

Keep your SDK and change the base URL. Route OpenAI, Anthropic, Google, DeepSeek, Qwen, Moonshot AI, and more through one governed path.

Get your API key Browse 500+ models

Start with one workload. Keep your provider accounts and SDK.

Your app or agent
Apertis ControlOne governed endpointkey · policy · route · usage
OpenAIAnthropicGoogleDeepSeekQwenMoonshot AI
model providers
30+
available models
500+
OpenAI-compatible surface
1

A gateway with operating context

Model access is the start. Control is the product.

Put compatible requests, access boundaries, routing decisions, and usage records in one managed path.

01

Keep the SDK you already use

Point an OpenAI-compatible client at Apertis, keep familiar request shapes, and select current model IDs from one catalog.

  • OpenAI-compatible endpoints
  • One API key
  • Current model catalog
02

Route with explicit control

Keep model choice visible while provider routes, fallback rules, and model policy govern where requests can run.

  • Provider routing
  • Fallback rules
  • Model policy
03

Bound access before workloads grow

Issue virtual keys, set quota boundaries, and separate agent or application access without distributing provider credentials.

  • Virtual keys
  • Quota controls
  • Governed access
04

See what consumed the budget

Inspect request, model, token, and cost records from one managed surface instead of reconciling separate provider consoles.

  • Usage logs
  • Model-level records
  • Budget visibility

Low-friction migration

One workload. Three steps.

Validate the compatible path and its usage record before expanding your rollout.

  1. 01

    Create one Apertis key

    Start with a personal key, then introduce bounded keys for separate agents, applications, or team workflows.

  2. 02

    Change the base URL

    Keep the OpenAI-compatible client and point its base URL at api.apertis.ai/v1.

  3. 03

    Choose a model and inspect usage

    Select a current catalog ID, send the first request, then review the request and spend record in Apertis.

PythonOpenAI-compatible
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_APERTIS_KEY",
    base_url="https://api.apertis.ai/v1",
)

response = client.chat.completions.create(
    model="YOUR_MODEL_ID",
    messages=[{"role": "user", "content": "Hello"}],
)

Before you switch

A clear first deployment path.

Is Apertis only a model proxy?

The compatible gateway is the execution surface. Apertis Control adds the managed access, routing, quota, policy, usage, and billing context needed to operate AI workloads across providers.

Can I keep using the OpenAI SDK?

Yes. Use the OpenAI-compatible endpoint, replace the base URL and API key, then choose an Apertis model ID from the catalog.

Do I need to move every workload at once?

No. Start with one agent or application, validate its model access and usage record, then move additional workloads when the operating path is clear.

How do I pay for usage?

Apertis supports Coding Plans and pay-as-you-go access. Review the current plan and model catalog before choosing the fit for your workload.

Start with one workload

Put the next model behind one governed API.

Create an Apertis key, switch one base URL, and verify the first request before expanding the rollout.

Get your API key Browse 500+ models