AI API Usage Tracking for Teams

Route API calls through Joule or sync from provider admin APIs. Track tokens, costs, and usage across OpenAI, Anthropic, and more—all in one dashboard.

100%

API Calls Tracked

Every request through the proxy is logged with tokens, cost, and latency.

7+

Providers Supported

OpenAI, Anthropic, Mistral, Cohere, Groq, Together, Perplexity, plus custom providers.

2

Admin API Syncs

Sync historical usage directly from OpenAI and Anthropic admin APIs.

5 min

Setup Time

Change your API endpoint to route through Joule. Works with existing SDKs.

What Joule Does

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

Proxy Gateway

Route AI API calls through Joule. We forward requests to providers and log tokens, cost, and latency automatically.

  • Works with existing OpenAI SDKs
  • Automatic logging on every request
  • Streaming response support
  • Add custom OpenAI-compatible providers
api-integration.ts

// Before: Multiple SDKs

import OpenAI from 'openai';

import Anthropic from '@anthropic-ai/sdk';

import { GoogleAI } from '@google/generative-ai';

// After: One unified API

import { Joule } from '@joule/sdk';

App
Your Code
J
Joule

Cost Tracking

See what you spend on AI. Track costs per request, user, team, and provider.

  • Real-time cost tracking
  • Historical spending trends
  • Per-request cost breakdown
  • Export data for reporting
Cost AnalysisLast 30 days

Monthly Savings

$1,247.80

Reduction

47%

Smart Routing Results

GPT-4 → GPT-4o-mini-$892
Claude 3 Opus → Sonnet-$284
Cached Responses-$71

AI Recommendation

Switch classification tasks to GPT-4o-mini for an additional $156/mo savings.

Usage Analytics

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

  • Token usage tracking
  • Request volume over time
  • Usage by team and user
  • Model usage breakdown
Analytics Overview

847K

Total Requests

+12%

234ms

Avg Latency

-8%

99.7%

Success Rate

+0.2%

Usage by Model

GPT-4o
42%
Claude 3.5
31%
GPT-4o-mini
18%
Other
9%

Team Management

Manage your organization with roles and permissions.

  • Role-based access (Owner, Admin, Member, Viewer)
  • Per-user usage tracking
  • API key management
  • Team invitations

Users Breakdown

Individual AI usage analytics

Total Users

26

Total Queries

12,847

Avg Queries/User

494

Top Users by Cost

sarah.chen
$286
mike.johnson
$165
emily.davis
$142
alex.kim
$98
jordan.lee
$72

All Users

UserRequestsCost
sarah.chen
Engineering
1,842$286
mike.johnson
Product
1,256$165
emily.davis
Research
982$143
alex.kim
Engineering
756$98
4 of 26 users

Provider Integrations

Track usage through proxy or sync from provider admin APIs.

  • OpenAI: Organization Usage API sync
  • Anthropic: Admin API + Claude Code Analytics
  • Proxy tracking: Mistral, Cohere, Groq, Together, Perplexity
  • Custom providers: Any OpenAI-compatible API
Connected ProvidersAll systems operational
OpenAI
GPT-4o, GPT-4o-mini, GPT-3.5

423K

requests

Anthropic
Claude 3.5, Claude 3 Opus

189K

requests

Google AI
Gemini Pro, Gemini Ultra

67K

requests

Auto-sync enabled
Last sync: 2 min ago

Claude Code Analytics

Track developer productivity metrics for teams using Claude Code.

  • Sessions, commits, and PRs by Claude Code
  • Lines of code added/removed
  • Tool acceptance rates (edit, write, notebook)
  • Per-developer breakdown
Budget & LimitsWithin Budget
Monthly Budget$2,847 / $5,000
57% used$2,153 remaining

Team Spending

Engineering$1420 / $2000
Product$892 / $1500
Research$535 / $1500

Rate Limits

1,000

Requests/min

2M

Tokens/day

50

Concurrent

100

Burst

Engineering approaching limit

71% of monthly budget used with 12 days remaining

Who Uses Joule

Teams that need visibility into their AI API spending.

Engineering Teams

Track API usage across your team. See who is using what, and how much it costs.

Finance & Ops

Get clear cost data for AI spending. Break down usage by team, project, or individual for budgeting and chargebacks.

Platform Teams

Centralize AI API access. Route all requests through one gateway with consistent logging and access control.

AI-First Companies

Track Claude Code usage alongside API costs. See developer productivity metrics and tool acceptance rates.

How It Works

Two ways to track usage: proxy your API calls or sync from provider admin APIs.

1

Connect Your Providers

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

2

Route Through Proxy or Sync

Either route API calls through Joule's proxy for real-time tracking, or sync historical data from provider admin APIs.

3

View Your Usage

See tokens, costs, and usage broken down by provider, model, team, and user in the dashboard.

Get Visibility Into Your AI Costs

Request access to start tracking your AI API usage.

Request Access