FAQ
Questions we hear before the first call
Still stuck on something? Send us a message and a human will answer.
What exactly is Lingyuns?
Lingyuns is a single API gateway in front of 60+ frontier and open language models, plus an on-demand GPU cloud. You fund one balance in USD and spend it on either tokens (metered per million) or GPU hours — no per-vendor accounts, no separate billing.
Do I need to change my code to switch?
No. The gateway is OpenAI-compatible: change
base_url to https://api.lingyuns.com/v1 and use your Lingyuns key. Everything else — messages format, streaming, tool calls, embeddings — stays the same.How does billing work?
Buy a token pack (2M – 135M credits) or a GPU reservation. US sales tax does not apply to our services; invoices are issued in USD. Your credits never expire while the account is active.
How do I pay?
After checkout we email you an invoice with bank-wire (USD) and USDT instructions. Once the transfer is confirmed, credits are activated and your API key is issued — typically within 2 business hours.
Are my prompts used for training?
No. We do not train on your data, and we operate a zero-retention mode for gateway requests. See the security page for details on encryption and logging.
Is there a free trial?
We issue trial keys for evaluation on request — write to [email protected] with your use case and expected volume.
Can I run dedicated GPU capacity?
Yes — everything from a single RTX 4090 to H200 clusters, hourly or monthly. For multi-node reservations talk to sales and we will size the fabric and storage for you.
What about uptime?
The gateway runs at a 99.9% monthly uptime target, with automatic failover between model providers. See our service level agreement.
Prefer to see it working?
Create a free account, browse the catalogue, and request a trial key for your evaluation.