FAQ

Quick answers to the most common questions about using Tchavi.

Is Tchavi OpenAI-compatible?

Yes. Tchavi is 100% compatible with the OpenAI API format. Just change the base URL and API key — the rest of your code stays the same. See Base URL.

How do I pay — what payment methods are supported?

You top up credits from your billing dashboard using mobile money. The full list of supported operators and countries is published live at /api/payments/providers. Credits are added instantly after payment. There is no subscription — Tchavi is pay-as-you-go. See Credits & Billing.

Do I get free credits to start?

There is no signup bonus — new accounts start with no credits. Your first top-up carries a +300 bonus credits on top of what you buy, and top-ups start at 1 000 FCFA.

Which models can I use before my first top-up?

Before any top-up your account is on the Free level: only Budget-group models are reachable (calling another model returns 403 plan_model_restricted) and requests are capped at 50 per day. Any top-up moves you to Builder and unlocks the whole catalogue. See Rate Limits.

How are credits calculated?

Each request consumes credits based on the model and the tokens processed. For chat models, input and output are billed per token at their own rates, and the total is rounded up once to a thousandth of a credit, with no minimum per request. Image models use a flat cost per image, TTS bills per 1K characters, transcription per minute of audio, and video per second of output. See Credits & Billing.

What models are available?

Tchavi gives you access to 80+ models across providers like OpenAI, Anthropic, and Google through a single API key. Browse them on the models page.

Are there rate limits?

Yes — two per-minute limits apply per user: RPM (requests per minute) and TPM (tokens per minute). Both depend on your account level, which is unlocked automatically by your lifetime spend. A global IP-based limit of 500 req/min also applies. See Rate Limits.

Is streaming supported?

Yes. Set stream: true to receive the response token-by-token as Server-Sent Events. See Streaming.

What happens if a video job fails — do I get a refund?

Yes. If a video job fails after billing (for example, an upstream error or timeout), the charged credits are automatically refunded.

Refunds also happen on success for reserve-billed models like kling-v3-motion-control: credits are held at the maximum possible output length when you submit, and the unused part is returned automatically once the actual length is known. The completed job shows the net charge in tchavi.credits_used and the refund in tchavi.credits_refunded.

How do I get an API key?

Create an account, then generate a key from the API keys page. The full key is shown once at creation, so copy it right away. See Authentication.