Total tokens tracked
Aggregated from teams and developers tracking AI usage across Claude Code and Codex. All data is private and anonymous.
Fastest acceleration
Pulling live public sync data.
Growth trajectory
Claude, Codex, and total tracked tokens
Works with the tools your team already uses
Your Team's AI Scoreboards — Visible Only to You
A private, secure dashboard that ranks your engineers across four dimensions of AI adoption. Only your organization can see this data — it never leaves your account.
Activity
Who is generating the most AI usage right now
Consistency
Who keeps usage steady day after day
Cache Efficiency
Who is creating the strongest savings signal
Adoption
Who is using AI tools across the broadest surface area
Private and secure by default
Scoreboards are scoped to your organization and protected by role-based access control. Only owners, admins, and authorized members can view rankings. No data is shared externally. This is not a public leaderboard — it's an internal tool for your team.
Know Who's Using AI — And Who Isn't
Per-engineer adoption tracking, activity heat maps, model preferences, and usage consistency over time.
Team Adoption
Last 30 days
Everything You Need to Measure, Manage, and Prove AI ROI
Built for engineering leaders who need visibility across their entire AI stack.
Budget Alerts
Get notified before costs exceed your limits
Cost Optimizer
Find cheaper models without sacrificing quality
Savings Engine
Compare subscription vs. pay-per-use plans
Data Export
CSV, JSON, or formatted reports for invoicing
Admin API Connections
Connect provider admin keys, sync automatically
MCP Server
One command connects Claude Code, Codex, and OpenCode
Wrapped Proxy Keys
Route API calls for attribution and spend limits
5-Minute Setup
Full visibility with no infrastructure changes
REST API
Access usage and cost data programmatically
Why Engineering Leaders Choose CostHawk
The visibility your team needs, without the overhead of building it yourself.
Get the answer before the board asks
Know exactly who on your team is using AI, how often, and whether adoption is growing or stalling.
See the full picture in one place
Every provider, every engineer, every project — unified in a single dashboard instead of scattered across billing portals.
Built for action, not just reporting
Budget alerts, cost optimization recommendations, and savings analysis so you can act on what you see.
No procurement headache
Free to start, no credit card required. Your team can be set up in under 5 minutes.
Start Free, Scale With Your Team
Free because every team deserves visibility into how they're using AI. No credit card required.
CostHawk
Everything you need to track and optimize your AI API usage
Team plans coming soon
Security-first by design
Whether you connect via Admin API keys, wrapped proxy keys, or MCP telemetry — your credentials are encrypted, we do not store prompt or code content, and you control what gets shared.
Enable auto-sync
One command installs the MCP and configures Claude + Codex auto-sync.
npm exec --yes costhawk@latest -- --login
AES-256 key encryption
Admin API keys are encrypted at rest with AES-256-GCM. We never store or see your keys in plaintext.
Wrapped key isolation
Proxy your API calls through CostHawk. Your real provider keys never touch your codebase or client devices.
Local-first parsing
MCP telemetry is computed locally from ~/.claude/projects and ~/.codex/sessions before anything leaves your machine.
No prompt storage
We only persist usage metadata: token counts, models, timestamps, and hashed project IDs. Prompt and code content is never stored.
Preview before upload
Dry-run syncs show exactly what would be sent, with payload previews and optional file lists for full transparency.
Opt-in auto-sync
Automatic uploads are off by default. Enable with COSTHAWK_AUTO_SYNC=true and disable Codex sync anytime.
FAQ