10 minutes

Quickstart

Sign up, copy the Costslice key, point your OpenAI client at us, tag three headers. Dollars show up on Tenants.

1. Create a workspace

   Sign in at costslice.com, name the org, copy cs_live_…. Put it in COSTSLICE_KEY. Keep OPENAI_API_KEY as the provider key.

2. Python openai

   from openai import OpenAI
   import os
   
   client = OpenAI(
       base_url="https://api.costslice.com/v1",
       api_key=os.environ["OPENAI_API_KEY"],
       default_headers={
           "x-cs-key": os.environ["COSTSLICE_KEY"],
           "x-cs-tenant": tenant_id,       # opaque id, never an email
           "x-cs-feature": "support-copilot",
           "x-cs-env": "prod",
       },
   )
   
   resp = client.chat.completions.create(
       model="gpt-5.6-terra",
       messages=[{"role": "user", "content": "Summarize this ticket"}],
   )

3. Node openai

   import OpenAI from "openai";
   
   const client = new OpenAI({
     baseURL: "https://api.costslice.com/v1",
     apiKey: process.env.OPENAI_API_KEY,
     defaultHeaders: {
       "x-cs-key": process.env.COSTSLICE_KEY,
       "x-cs-tenant": tenantId,
       "x-cs-feature": "support-copilot",
       "x-cs-env": "prod",
     },
   });
   
   const resp = await client.chat.completions.create({
     model: "gpt-5.6-terra",
     messages: [{ role: "user", content: "Summarize this ticket" }],
   });

4. Vercel AI SDK

   import { createOpenAI } from "@ai-sdk/openai";
   import { generateText } from "ai";
   
   const openai = createOpenAI({
     baseURL: "https://api.costslice.com/v1",
     apiKey: process.env.OPENAI_API_KEY,
     headers: {
       "x-cs-key": process.env.COSTSLICE_KEY!,
       "x-cs-tenant": tenantId,
       "x-cs-feature": "support-copilot",
       "x-cs-env": "prod",
     },
   });
   
   const { text } = await generateText({
     model: openai("gpt-5.6-terra"),
     prompt: "Summarize this ticket",
   });

   If tenant/feature change per request, pass headers on the call instead of the client. The three Costslice headers are strings, max 128 characters. Streaming works; we read usage on the final chunk (we set stream_options.include_usage when you stream). If usage is missing we estimate tokens and mark cost_source=estimate. Unknown models cost $0 and show as unpriced.

Local proxy

Point base_url at http://127.0.0.1:8788/v1 and keep the same headers. See the README for Next.js + Worker + Postgres.