What we do
ShinLinkAI is an independently operated AI API gateway. We aggregate the major model families — DeepSeek from DeepSeek, Alibaba's Qwen, Moonshot AI's Kimi, and GLM — behind one API key, one balance and two industry-standard protocols: Chat Completions and Messages.
In practice you change one line of configuration (the base URL) and your existing tools keep working. Cursor, Cherry Studio, mainstream IDEs, VS Code extensions and any official SDK can all point at the same gateway and share the same prepaid balance.
Why we exist
Using several model providers today is needless pain:
- Accounts pile up.Each one means separate signup, separate verification, separate dashboard.
- Payment gets annoying.Overseas cards get declined, minimum top-ups are high, currencies differ.
- Bills are scattered.Usage is spread across portals that do not talk to each other.
- Switching costs are real.Every new model means a new SDK, a new endpoint, a new failure mode.
ShinLinkAI pulls this together into one key, one balance and one compatible API — so teams spend their time building products instead of maintaining vendor pipelines.
Our pricing
Every model is billed at the vendor's publishedofficial list pricewith no mark-up-then-discount promotions. Every cent you pay corresponds to the vendor's published rate —prices are stable over timeand do not move with campaigns.
Prepaid balance,never expires. No subscription, no seat fees, no monthly minimum. From ¥100, topped up by scanning a WeChat Pay or Alipay QR code (bank cards can be linked; no overseas credit card required); each request is metered at the published per-model rate, with input, output, cache reads and reasoning tokens priced separately.
Everything is transparent: model prices, token-level usage, per-key limits and request history are all visible in the console. If any number does not add up, you can trace it to the exact request.
How the technology runs
Requests hit our edge nodes, are normalized to one protocol, and are routed to the fastest healthy upstream for the target model, then streamed back as standard SSE. Every customer's keys are isolated, with independent rate limits and spend controls; upstream failures are retried automatically before they reach your application.
This architecture is also why one key can serve hobby projects, production agent clusters and a reseller's customer base at the same time without interfering — each key has its own quota, limits and usage record.
Who uses ShinLinkAI
Independence and compliance
ShinLinkAI is an independent service: it is not affiliated with, endorsed by, or an official channel of any model vendor. Model names are used only to describe API compatibility. If you build a product on top of us, please make the same clear to your users.
ShinLinkAI is operated by Shanghai Xinrui Industrial Co., Ltd. (沪ICP备14051630号-1); the filing and payment entities are the same. Business customers can request VAT invoices (issued for the actual amount spent, with the invoice title matching the verified account entity).
We ask the same of all users: no abuse, no fraud, no false advertising. We reserve the right to suspend accounts and keys that harm the service or other customers.
Support
Support is provided by real people, in Chinese and English. Email support@shinloo.com or email support@shinloo.com. There is also an AI support assistant in the bottom-right of every page for instant questions about pricing, models and top-ups.
Where the name comes from
"ShinLinkAI" carries the sense of "speed and summit" — the strongest model, one lightning bolt away. The product is deliberately plain: clear prices, clear limits, no hidden rules — infrastructure is at its best when it is "boring", and that is exactly what we mean.