Introducing the CloudService blog
A short note on what we'll publish here: engineering notes, billing explainers, and product updates from the CloudService team.
What this blog is for
CloudService gives you prepaid access to leading AI models through one API key. This blog is where we'll write about the parts worth explaining.
Expect three kinds of post:
- Engineering notes — how the gateway actually works, and what we learned building it
- Billing explainers — credit vs token modes, and why the distinction matters
- Product updates — new models, new surfaces, changes worth knowing about
Using the API
Every model is reachable through one OpenAI-compatible endpoint:
bash
curl https://api.cloudservice.services/v1/chat/completions \
-H "Authorization: Bearer $CLOUDSERVICE_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"gpt-5.4-mini","messages":[{"role":"user","content":"Hello"}]}'
Claude models are also available on a native Anthropic surface at /v1/messages, so existing SDKs work without changes.
Model availability
| Family | Example | Wire format |
|---|---|---|
| GPT | gpt-5.4-mini | OpenAI-compatible |
| Claude | claude-sonnet-5 | Native Anthropic |
Prepaid means no monthly commitment. You buy credit, you spend credit, and unused balance stays yours.
More soon.