Track Your AI API Costs

Route API calls through Joule or sync from provider admin APIs. See tokens, costs, and usage across OpenAI, Anthropic, and more.

Total Cost

$127.84

4,521 requests

Total Tokens

2.4M

Avg 531 per request

Avg Response Time

892ms

Time to complete

Energy Used

12.4 Wh

Total watt-hours

Carbon Emissions

8.2g

CO2 equivalent

Water Usage

45.3 mL

For cooling

Cost Over Time

Cost ($)

Token Usage

Tokens

Response Time

Avg Response Time (ms)

Model Usage

62565554
GPT-4o
Gemini Pro
Claude 3.5
GPT-4o-mini

What Joule Does

Track AI API usage across your organization. See what you're spending and where.

Proxy Gateway

Route API calls through Joule. Every request is logged with tokens, cost, and latency.

Cost Tracking

See exactly what you spend on AI. Breakdown by provider, model, user, and team.

Usage Analytics

Dashboard showing tokens, requests, and costs across all your AI providers.

Team Management

Invite members, assign roles, and track usage per person.

Provider Sync

Sync historical usage from OpenAI and Anthropic admin APIs. More providers coming.

Claude Code Analytics

Track developer productivity: sessions, code changes, commits, and tool usage.

How Joule Works

Simple integration, complete visibility. Route your API calls through Joule and we handle the rest.

1

Connect Your Providers

Add your API keys for OpenAI, Anthropic, or any provider. Keys are encrypted and stored securely.

2

Route Through Joule

Update your API endpoint to use Joule's proxy. We forward requests to your provider and track everything.

3

Get Full Visibility

See costs, tokens, latency, and usage per user, team, or project in real-time.

Simple endpoint change
// Before: Direct to OpenAI
const endpoint = 'https://api.openai.com/v1/chat/completions'

// After: Through Joule proxy
const endpoint = 'https://api.joule.ai/proxy/openai/chat/completions'
// Add headers: x-joule-key, x-provider-key

SDKs for Every Language

Wrap your existing OpenAI or Anthropic SDK with Joule. Usage is tracked automatically with zero code changes to your API calls.

  • TypeScript, Python, and C# SDKs
  • Wraps existing OpenAI and Anthropic SDKs
  • Automatic token and cost tracking
  • Supports streaming responses
app.ts
import OpenAI from 'openai';
import { JouleOpenAI } from '@joule/sdk';

const openai = new OpenAI({
  apiKey: process.env.OPENAI_API_KEY
});

// Wrap with Joule for automatic tracking
const joule = new JouleOpenAI(openai, {
  apiKey: process.env.JOULE_API_KEY
});

// Use exactly like the normal SDK
const response = await joule.chat.completions.create({
  model: "gpt-4",
  messages: [{ role: "user", content: "Hello!" }]
});
// Usage automatically tracked to Joule
Install:npm install @joule/sdk

Cost Optimization Recommendations

Based on your usage patterns, we found opportunities to save money.

Potential Monthly Savings

$284.50

Recommendations

3

Model Recommendation

HIGH CONFIDENCE

Current Model

gpt-4o

Recommended Model

gpt-4o-mini

Estimated Monthly Savings$142.80
Cost Reduction58.2%

For simple classification tasks, gpt-4o-mini provides similar accuracy at a fraction of the cost.

Trade-offs:

  • Slightly lower reasoning capability for complex tasks
  • May require prompt adjustments for edge cases

Clear Cost Visibility

See exactly where your AI budget goes. Break down costs by provider, model, team, and individual user.

  • Real-time cost tracking per request
  • Historical usage trends and charts
  • Per-user and per-team breakdowns
  • Export data for reports

Team Management Built-In

Manage your entire organization's AI access from one dashboard. Invite team members, assign roles, and control permissions with ease.

  • Role-based access control (Owner, Admin, Member, Viewer)
  • Team invitations and member management
  • Per-user analytics and cost tracking
  • Team-level spending insights

Team Members

Manage your organization's access

JD
John DavisOWNER

john@company.com

Last active

Just now

SK
Sarah KimADMIN

sarah@company.com

MP
Mike PetersMEMBER

mike@company.com

alex@company.comPENDING

Invited as Member

AI Provider Connections

Connect your AI provider accounts

OpenAI

GPT-4, GPT-3.5

Synced today
Anthropic

Claude 3.5, 3 Opus

Synced today
Google AI

Gemini Pro, Ultra

Synced yesterday
Internal LLM

Custom API

Custom
llm.internal.company.com

Supported providers include:

MistralGroqCohereReplicateHugging Face+ Custom APIs

Supported Providers

Track usage through the proxy or sync historical data from provider admin APIs.

Full API Sync

OpenAI (Organization API) and Anthropic (Admin API with Claude Code Analytics) — sync historical usage directly from provider.

Proxy Tracking

Mistral, Cohere, Groq, Together AI, Perplexity — track API calls through the proxy.

Custom Providers

Add any OpenAI-compatible endpoint — self-hosted models, internal APIs, or other providers.

Know what you're spending on AI

Get visibility into your AI API usage. Request early access.

Request Access