Relay services that cover several vendor families at once — Claude, GPT, Gemini, Grok and more. Aimed at developers who mix models inside one project and would rather maintain a single integration.
Vitas mainly provides independent key pools for GPT enterprises, with SLA guarantee of 99% for u…
Multi-model one-stop aggregation: GPT, Claude, Gemini, Deepseek, GLM, Kimi, and video. All major…
LinoRoute aggregates 500+ models from domestic and overseas sources through third-party upstream…
The platform only offers official API keys and periodically provides discounted official models.
Source of National Models Self-constructed ID Pool Model Fidelity Verifiable License
Per the service's own site: one of the earlier relays, and it says its long-run stability has st…
TokenLab is an OpenAI-compatible model gateway. One API key lets you call 380+ AI models for tex…
There is not enough measured data in this category to justify a recommendation — either few services are listed here, or we have not yet collected enough uptime and latency samples. We do not invent recommendations in that situation: rather than give you a conclusion without evidence, we would rather say plainly that we have none yet.
You can compare the services yourself in the table below, or browse every service in this category sorted by model coverage. Once enough samples accumulate, a recommendation will appear here automatically.
This page shows you what exists; the pages below show you who is more stable, faster and cheaper. Read together, they tell you more than either alone.