Docs

Two-line migration

Point any OpenAI-compatible client at Sluice and set the model to auto.

TypeScript

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://sluice.forum/api/v1",
  apiKey: process.env.SLUICE_KEY, // sk-sluice-… from the dashboard
});

const answer = await client.chat.completions.create({
  model: "auto", // or any OpenRouter model id to skip routing
  messages: [{ role: "user", content: "Hello" }],
});

Create a key in the dashboard under Gates. Each key spends your Orbio balance up to its daily and monthly limits. Streaming works with stream: true; streamed answers skip the answer check and retry.

Every response carries x-sluice-model, x-sluice-tier, x-sluice-cost-usd and x-sluice-saved-usd. Agents: see the agent guide or llms.txt.

Routing controls

model: "auto" lets Sluice pick.
Any other model id is passed through unchanged.
x-router: off skips routing for one request.

Tiers right now

Tier	Models, in config order
easy	anthropic/claude-haiku-5.5, deepseek/deepseek-v4-pro, google/gemini-3.5-flash-lite
medium	anthropic/claude-sonnet-5.5, openai/gpt-5.6-sol, x-ai/grok-4.7
hard	anthropic/claude-opus-5.5, anthropic/claude-fable-5.1

Within a tier Sluice picks the cheapest model for the request. Savings are measured against anthropic/claude-fable-5.1 at list price.

Privacy

Sluice stores billing metadata only: model, tokens, cost, latency. Prompts and answers are never stored. Accounts with Orbio Incognito on are not supported yet; their requests are refused rather than sent as plaintext.