SilvR AI
Get API key
Get API key

One gateway, every model

One base URL.
Every model.

Keep the SDK you already use. Point it at our base URL and call any model in the catalog.

Get an API key See the live catalog
One API keyOne balanceNo new SDK

One base URL

Change one line.

Keep your SDK, streaming, and tool calls. Change the base URL and the key.

api.silvrai.net ready
POST/v1/chat/completions
import OpenAI from "openai";const ai = new OpenAI({  apiKey: process.env.SILVR_API_KEY,  baseURL: "https://api.silvrai.net/v1",});const response = await ai.chat.completions.create({  model: "gemini-2.5-flash",  messages: [{ role: "user", content: "Explain this trace." }],});
Chat Completions or Responses matches your client
Every request above reaches the same model. Responses come back the way your client expects.

Catalog

Every model, one endpoint.

One key, one balance, one base URL. New models land as providers ship them.

MODEL FAMILIES
OpenAI Anthropic Google Gemini DeepSeekand more
Live catalog

What we run

We manage the upstream.

Provider accounts, rotation, retries, rate limits, model and price updates. All on our side. Yours is one endpoint and one key.

  • One account and one key for the whole catalog.
  • Providers rotate and retry. Your request just returns.
  • New models land in the catalog as providers ship them.
01 / ROTATION

No single upstream.

Providers run in rotation. A rate-limited or failing one drops out and the next takes over.

02 / ANY PAIRING

Client and model are separate.

The SDK you already use can call any model in the catalog. They don't have to match.

03 / DROP-IN

The base URL is the whole diff.

Standard auth headers, SSE streaming, tool calls. Nothing else in your stack moves.

What you pay

A fraction of the provider's own price.

We meter what each request used — tokens, seconds, images, characters — at the model's current price, then apply your plan's share.

WHAT LEAVES YOUR BALANCE
usage × model's current price×your plan's share
Metered on reported usageCharged after the response
See every model's upstream pricing
METERED USAGEreq_7f2a
SETTLED
gemini-2.5-flash2,300 tokens
Input
1,220
Cache read
600
Output + reasoning
480
Cost
$0.0055
Plan rate
×0.95
Total$0.0052
PLANS

Included usage each period. Your balance covers the rest.

freeFree to start$3.5/day included
Lite$10 / month$30/day included0.95× costwe cover 5%
Regular$25 / month$80/day included
Cuties$40 / month$180/day included

Questions

Straight answers.

Keys, pricing, and the basics.

01Do I need a new SDK?

No. Keep the SDK you already use. Point it at our base URL and drop in your SilvR key.

02Which models can I call?

Everything in the live catalog. New models land as providers ship them.

03What happens when my balance runs out?

Requests are refused until you top up or buy a plan.

04How do I pay?

Crypto, via the dashboard. Your balance is credit you spend across any model.

Start here

Change the base URL. That's it.

One key, one balance, every model.

Create an account See the code