Attribution

Attribute AI spend to customers.

Tag a call once. Every request after that arrives priced in dollars and already labeled.

The same AI spend.
Now in your business terms.

Without SuperPenguin · By modelkimi-k2p7-code$3,146gemini-3.1-pro$2,742gpt-5.2$2,580claude-sonnet-4.5$2,568gpt-5.2-mini$1,637gemini-3.6-flash$1,496claude-haiku-4.5$1,226text-embedding-3-large$263$15,658

With SuperPenguin · By Customer

+ Any tag

Ridgepoint$8,696Globex$4,460Initech$2,502$15,658

Connect every request to your business.

Wrap your existing provider client, then attach the customer, feature, team, or any metadata you already use.

CustomerRidgepoint
FeatureSupport agent
TeamCX
EnvironmentProduction
PlanEnterprise

Use the cost that matches your bill

Each request is priced using provider-reported usage. You can instead set your own negotiated or private model rate, too.

Usage

10K

tokens

Price

per 1M

Cost

$0.02

this request

Catch the surprise before it becomes the bill

Get a Slack, Discord, or email message when spend crosses a threshold, before the invoice surprises you.

# ai-cost-alerts

Slack

SuperPenguin

9:42 AM

Support agent spend is 4.3× higher

$412.36 today

See what changed

Two lines and you're tracking

pip install superpenguin # or: npm install @superpenguin/js

your_app.py

import superpenguin as sp
from openai import OpenAI

sp.init(api_key="sp-...")

client = sp.wrap(OpenAI(), metadata={
    "customer_id":     "ridgepoint",
    "feature":         "support_agent",
    "team":            "cx",
    "environment":     "production",
    "prompt_key":      "ticket_summary",
    "prompt_version":  "3",
})

# Calls go directly to OpenAI. Cost is tracked automatically.
response = client.chat.completions.create(
    model="gpt-5.4",
    messages=[{"role": "user", "content": prompt}]
)