FAQ

How is OpusGate cheaper than going direct?

We aggregate traffic and resell tokens. You get the exact same models at up to 10× lower cost than the official list price.

Which models can I use?

Claude (Opus 4.8/4.7/4.6, Sonnet 4.6, Haiku 4.5), GPT 5.4/5.5 and Qwen 3.7 — the original vendor models, unchanged. We do not fine-tune, filter or modify responses.

Do you store my prompts?

We log request metadata (timestamp, model, token counts, cost) for billing and abuse prevention. Prompt and completion contents are not stored on OpusGate servers.

What payment methods do you accept?

Card (via Paddle) and crypto (via CryptoCloud) — both self-serve. Minimum top-up is $2. Bonuses: $10 → +5%, $50 → +7%, $100 → +10%.

Refunds?

Unused balance is refundable on request within 14 days of the top-up, minus payment-processor fees. Used credits are non-refundable.

Which API protocols are supported?

Anthropic Messages API (POST /v1/messages) and OpenAI Chat Completions (POST /v1/chat/completions). Streaming (SSE) works for both.

Are there rate limits?

No artificial per-minute limits. Concurrency is bounded only by what the upstream model can serve.

Where are you based?

European Union. Infrastructure is in EU data centres.

How do I get support?

Email support@opusgate.dev or use the in-app support chat.