AI built into your apps, with a cap on every key

Your apps and your team use Claude, Gemini, Llama and many more models through one CloudWish key. Give each app its own key and a monthly spend cap, and see AI usage as its own line on your bill.

Early access. Prices are published when it opens.

One key per app, one cap per key

Every app that needs AI gets its own CloudWish key. You decide how much each one may spend in a month, and you see what each one used.

  • Your n8n workflows, your AI agent and your website each get a separate key
  • Set a monthly spend cap per key, and change it any time
  • AI usage shows as its own line on your CloudWish bill
  • Developers keep the OpenAI request format they already use
Preview: request format
curl https://ai.cloudwish.com/v1/chat/completions \
  -H "Authorization: Bearer $CLOUDWISH_KEY" \
  -d '{
    "model": "route:cheapest",
    "fallback": ["claude", "gemini"],
    "messages": [
      {"role": "user", "content": "Summarise this ticket"}
    ]
  }'
route:cheapestdeepseek

Many models, one key

The CloudWish AI router sends each request to the model you name, or picks one by a rule you choose. Switching models is a change to one line, not a new account.

  • Name a model

    Ask for Claude, Gemini or any other model by name when you need that one.

  • Or route by rule

    Let each request go to the cheapest model that fits, or the fastest.

  • Fallback built in

    If a provider fails, the request retries on another instead of returning an error to your app.

Already connected to your apps

AI apps from the catalogue arrive with their key in place and the cap you set. OpenClaw is ready at launch.

Browse all AI apps

AI questions

Which models can I use?

Claude, Gemini, Llama, Mistral, DeepSeek, Qwen, Perplexity and more, all through one CloudWish key.

How do spend caps work?

Each key has a monthly cap that you set. A key does not spend past its cap, and you can raise or lower it at any time.

Do I need to change my code?

Very little. The API uses the OpenAI request format, so most apps only need the CloudWish address and key.

How is AI billed?

AI usage is its own line on your CloudWish bill, next to your plans and servers, broken down per key.

Can my VPS apps use it?

Yes. Apps on a CloudWish VPS get their key set up for them. See VPS.

When does it open?

It is in early access. Join the list at the top of this page and we email you when it opens.

Put AI in your apps, with a cap you set.