TICON AI is a prepaid AI API. You top up a wallet in Bangladeshi taka (BDT), create an API key, and call a gateway that speaks the OpenAI and Anthropic wire formats. Each request is metered by tokens and debited from the wallet.
Base URL and endpoints
Every endpoint lives under one base URL:
text
https://api.ticonai.com/v1| Method | Path | What it does |
|---|---|---|
| POST | /v1/chat/completions | OpenAI-compatible chat completions, with optional streaming. |
| POST | /v1/messages | Anthropic-compatible messages, with optional streaming. |
| GET | /v1/models | The models your key can call. |
| GET | /v1/models/{id} | One model. |
| GET | /healthz and /readyz | Liveness and readiness (no key needed, served at the API origin, not under /v1). |
Both chat dialects are translated to one internal request, so billing and routing behave the same whichever you use.
Check that the service is reachable (no key and no cost):
bash
curl -sS https://api.ticonai.com/healthzHow a request flows
- You send a request with your API key.
- The gateway checks the key, its restrictions, the rate limits and your wallet.
- It reserves a hold on the wallet, the most the request could cost.
- The request runs. When it ends, the hold is settled to the real cost, or released if nothing was produced.
- You receive the response, with an
x-request-idheader you can quote to support.
Read Billing and wallet for the exact rules.
Where to start
- Quickstart: your first request in a few minutes.
- Authentication and API keys.
- Chat completions and Messages.
- Streaming, Models, Errors, Rate limits and Idempotency.
- Best practices and SDK examples.
Product changes are listed in the changelog.