Apertis for Kilo Code / Kilo Gateway

One OpenAI-compatible endpoint across all major providers. Predictable Coding Plan capacity. PAYG overflow with a hard cap. apertis.ai/log shows exactly which key, model, and request burned how much. For teams looking at a paid pilot with budget control and per-agent spend visibility, talk to us — Apertis Team Gateway is a manual pilot today, not a self-serve product.

Who this is for

• Developers using the Kilo Code VS Code extension
• Self-hosted Kilo Gateway operators
• Small teams (2–5 people) who want one billing surface and a single audit trail across multiple coding agents

Prerequisites

• Kilo Code or Kilo Gateway installed — see the Kilo Code docs for the latest install method
• An Apertis account and API key (apertis.ai/settings/apiKeys)
• For team pilots: contact us to configure the pilot key, monthly spend cap, and reporting cadence

Setup

1. Pick a Coding Plan

For solo Kilo use, Pro ($25/mo) or Plus ($60/mo) usually fits. For team pilots, the spend cap and reporting cadence are agreed in writing before the pilot key is issued — see the Team Gateway section below.

2. Create an API key

apertis.ai/settings/apiKeys → Create API Key. Name it something you'll recognize later, e.g. kilo-vscode or kilo-gateway-prod.

3. Point Kilo at Apertis

Apertis exposes an OpenAI-compatible endpoint at https://api.stima.tech/v1. Configure Kilo the same way you would point it at any OpenAI-compatible provider — Kilo's settings UI exposes a custom-endpoint field; see the Kilo Code docs for the exact location in your installed version.

The two values Kilo needs:

Base URL: https://api.stima.tech/v1
API Key:  sk-YOUR_APERTIS_KEY

For Kilo Gateway (self-hosted), add Apertis as an upstream OpenAI-compatible provider in your gateway config — refer to the Gateway's documented provider schema for the exact YAML / JSON shape.

4. Verify

Open any code file and trigger a Kilo action. A successful generation means the integration is working. From the gateway, a simple smoke test:

curl -sS https://api.stima.tech/v1/chat/completions \
  -H "Authorization: Bearer sk-YOUR_APERTIS_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.5","messages":[{"role":"user","content":"Reply with OK"}]}' \
  | head -50

5. Monitor usage

apertis.ai/log filters by key, by model, and by time range. For team pilots, the operator pulls usage exports during the pilot until the team dashboard ships.

Team Gateway pilot

Apertis Team Gateway is currently a manually-onboarded paid pilot. Self-serve team controls (per-member quota, hard spend caps enforced at the gateway, per-project keys) are on the roadmap but not live. During the pilot phase you get:

• A dedicated API key with an agreed monthly spend cap
• A written agreement on the cap and reporting cadence
• Weekly or monthly usage exports until automated reporting ships
• Direct support during onboarding

Talk to us about Team Gateway

Choosing a model

Current full list and pricing: apertis.ai/models. Apertis routes a single model name to whichever upstream provider has capacity, so the catalog moves faster than any static list in this guide.

Common picks for Kilo workflows:

• Team default: gpt-5.5, claude-sonnet-4-6
• Complex reasoning / architecture review: claude-opus-4-7
• Cost-optimized variant for high-frequency users: glm-5.1, deepseek-v4-pro, minimax-m2.7

PAYG Overflow

Every plan includes optional pay-as-you-go overflow. Once you exhaust the monthly Coding Plan quota, Apertis charges per-request at the underlying model's cost. You can cap monthly PAYG spending at apertis.ai/settings/billing.

For team pilots, PAYG overflow is opt-in and capped per the written pilot agreement — it is never enabled silently.

Common errors

Start here

Create your Apertis API key