BurnRate requires JavaScript to display interactive analytics and charts. Please enable JavaScript in your browser to use the full experience.

100% free to start — no credit card required

See everything your AI agents do — and what they cost.

Free local-first observability for Claude Code, Cursor, Copilot, and 4 more providers. Track costs, observe subagent behavior, and optimize spending — all from your terminal. Zero config, zero account, zero cloud dependency.

brew install --cask burnrate-dev/tap/burnrate && burnrate⎘

Download Binary

Completely free for individual developers — no signup needed. Your code never leaves your machine.

Planning a team budget? Estimate AI-tool spend free →

BurnRate

See exactly what your AI coding tools cost you

Check it out on Product Hunt →

BurnRate Dashboard — Overview showing plan value, cost tracking, and daily token usage

A dashboard built for developers

Nine analytics views to observe your AI agents and understand spending from every angle.

Cost Breakdown & Plan Value

See exactly what your AI usage would cost on API pricing versus what you pay on your subscription. Track daily spending trends and model-by-model cost distribution.

Daily API equivalent cost chart
Model cost breakdown (donut chart)
Plan value multiplier calculation
Savings percentage vs API pricing

Cost breakdown with daily trends and model distribution

Session Explorer

Browse every AI session with full metadata — duration, tokens, cost, model, and project. Filter by date, project, or model. Sort by any column to find your most expensive sessions.

Searchable session history
Per-session API equivalent cost
Filter by project, model, date range
Cache hit rate per session

Session explorer with sortable columns and filters

Provider Comparison

Side-by-side comparison across all 7 providers. Per-provider model breakdowns, efficiency scores, token distribution, and daily cost sparklines.

Cost-per-session and cost-per-message metrics
Efficiency scoring (0-100)
Per-model cost breakdown within each provider
Daily cost trend sparklines

Provider comparison with model breakdowns and efficiency scores

Optimization Engine

46 rules analyze your usage patterns and suggest concrete, actionable savings — with provider-specific config snippets for each tool. Each rule explains why it matters, how much you can save, and provides one-click config snippets to apply.

Impact scoring (high/medium/low)
Estimated monthly savings per rule
Auto-apply config snippets to CLAUDE.md
Educational explanations with real-world analogies

Optimization engine with actionable savings recommendations

Project Analytics

Track costs and usage per project. See which codebases consume the most tokens, set per-project budgets, and identify projects that need optimization.

Per-project cost and session counts
Project cost ranking charts
Budget cap monitoring
Cross-project comparison

Project-level cost analytics and rankings

Usage Patterns & Operations

Understand which operations — reading, writing, searching, thinking — consume the most tokens. Identify tool usage patterns and optimize your workflow.

Operation-level token breakdown
Tool call frequency analysis
Context efficiency metrics
Session depth distribution

Usage patterns showing operation costs and tool analytics

Rate Limit Monitoring

See exactly how close you are to hitting rate limits across Claude, Cursor, and Codex. Live utilization bars with color-coded thresholds and reset countdowns.

5-hour, daily, and weekly window tracking
Color-coded utilization (green → yellow → red)
Reset time countdowns
Multi-provider limit tracking

Rate limit monitoring with utilization bars and reset timers

Budget Alerts

Set per-project spending caps and burn rate limits. Get desktop notifications when budgets are at risk. Configure Slack webhooks for team-wide alerts.

Per-project monthly and daily caps
Burn rate monitoring and projections
Desktop and Slack notifications
Budget vs actual visualizations

Budget management with spending caps and alert configuration

Works with Claude Code Cursor IDE OpenAI Codex GitHub Copilot Windsurf Cline Aider

Optimization Rules

40+

AI Models Tracked

Providers Supported

Real-time

Budget Alerts

Agent observability meets cost analytics

Powerful analytics that run entirely on your machine. See what your agents do, what they cost, and how to optimize them.

Subagent Observability

See every subagent Claude Code spawns — what it costs, what tools it uses, how long it runs. Per-subagent cost trees show exactly where tokens go in agentic workflows. The only tool that provides this — Anthropic closed feature requests #10164 and #10388.

Per-subagent cost breakdown
Tool call tracking per agent
Agent duration and depth analysis
Zero config — parses local JSONL automatically

Multi-Provider Analytics

Track Claude Code, Cursor, Codex, Copilot, Windsurf, Cline, and Aider in one dashboard. Compare costs across all 7 providers.

Cost Optimization Engine

46 optimization rules analyze your usage patterns and surface concrete opportunities — including provider-specific rules for Claude Code, Cursor, Copilot, Codex, Windsurf, Cline, and Aider. Your result depends on actual tools and usage.

Model Recommendations

AI-powered suggestions for which model fits each project. Stop overpaying for simple tasks.

Operation Cost Breakdown

See exactly which operations -- reading, writing, searching, thinking -- consume the most tokens.

Context Efficiency Analysis

Understand how your context window fills up across sessions. Identify context bloat patterns.

Local-First & Private

Your code and raw session data never leave your machine. Core analytics have no cloud dependency.

Set Budgets. Get Alerts. Stay in Control.

Set per-project spending caps, burn rate limits, and alert thresholds. Get desktop notifications or Slack webhooks when budgets are at risk.

Per-project monthly and daily caps
Burn rate monitoring
Desktop notifications
Slack and email alerts for teams

Plan Usage Limits & Rate Tracking

See exactly how close you are to hitting rate limits across Claude, Cursor, and Codex. Live utilization bars with reset countdowns — never get throttled by surprise.

Claude 5h / 7-day / Opus window tracking
Color-coded utilization (green → yellow → red)
Reset time countdowns
Cursor & Codex activity monitoring

Team-Wide AI Cost Intelligence

See how your whole team uses AI tools. Identify high-spenders, optimize provider mix, and set budgets that keep everyone productive.

Per-member cost tracking
Team budget management
Provider consolidation insights
Optimization leaderboards

Three steps. Zero config.

Get full agent observability and cost visibility in under a minute.

Install

brew install --cask burnrate-dev/tap/burnrate && burnrate
Install and launch in one command.

Analyze

Parses your local AI tool data.
Nothing uploaded, ever.

Optimize

Get actionable recommendations
to reduce token waste.

AI coding tools are expensive.
Are you getting your money's worth?

$100/mo

Claude Max

"But how much of that capacity are you actually using?"

$20/mo

Cursor Pro

"Which features drive the most token usage?"

Pay-per-token

OpenAI API

"Do you know your actual cost per coding session?"

The only AI agent observability platform

Anthropic and OpenAI show you a billing page. We show you where every token goes — down to each subagent.

Multi-provider tracking

✓

✗

Subagent observability

✓

✗

CLI/IDE tool tracking

✓

✗

Per-project cost breakdown

✓

✗

Optimization recommendations

✓

✗

Model recommendations

✓

✗

Team analytics

✓

✗

Data stays local

✓

N/A

Built for Engineering Teams

Give engineering leaders visibility into AI tool ROI without compromising developer privacy

Team-Wide Cost Visibility

See aggregate AI spending across your entire engineering org. Track costs per team, per project, per developer -- without accessing their code.

ROI Measurement

Know if your $100/seat Claude Max licenses are actually being used. Compare actual usage vs subscription costs across the team.

Standardization Insights

Discover which AI tools your team actually uses. Make data-driven decisions about which providers to standardize on.

Privacy-First Architecture

Raw code and session data stay on developer machines. Team sync sends only aggregated usage and cost metrics when enabled.

How Teams Use It

From install to insights in minutes, not weeks.

Admin creates team

Set up your team on getburnrate.io and invite developers

Devs install locally

Each developer installs the single binary and enters their team license key

Metrics sync automatically

Aggregated cost data syncs to the team dashboard. No code leaves the machine.

Optimize team spending

See team-wide analytics, identify optimization opportunities, track ROI

Ready for a team-wide cost view?

Start locally for free, then compare team plans when you need aggregate usage and budget visibility.

Compare Team Plans

Built for developers who care about agent visibility and AI spend

Subagent Trees

7+ Providers

40+ AI Models

46 Optimization Rules

100% Local & Private

Enterprise

Everything your organization needs to manage AI tool spending at scale.

SSO / SAML

Dedicated support

Custom integrations

SLA guarantee

Volume discounts

On-premise option

Talk to Sales

Simple, transparent pricing

Start completely free — no account or credit card needed. Upgrade only when you want deeper insights.

Free

forever

Overview, Sessions, Costs tabs
7+ providers (Claude, Cursor, Copilot, Windsurf, Aider, Cline, Codex)
30-day history
Basic analytics dashboard

Get Started Free

Pro

$12/month

Save $50+/mo on average

Everything in Free
46 optimization rules (provider-specific)
Budget controls and notifications
Provider comparison dashboard
Model recommendations
Operation cost breakdown
Context efficiency analysis
Export to JSON/CSV

Start Pro

Best for Teams

Team

$29/seat/mo

For engineering teams

Everything in Pro
Team dashboards
Team budget management
Slack and webhook alerts
Per-member cost tracking
Priority support

Per seat, per month. Includes team dashboard, aggregate analytics, and admin console.

Start Team

Enterprise

Custom

For teams of 50+

Everything in Team
SSO / SAML
Dedicated support
Custom integrations
SLA guarantee
Volume discounts

Contact Sales

Frequently asked questions

Does my data leave my machine?

Raw code, prompts, and session content stay local. BurnRate sends a one-time anonymous install ping with version and platform metadata, which you can disable with DO_NOT_TRACK=1. Pro and Team also validate license keys; team sync sends aggregate metrics only when enabled.

What is subagent observability?

When Claude Code runs complex tasks, it spawns subagents — child processes that each consume tokens independently. BurnRate parses your local JSONL session data and builds a tree view of every agent and subagent: what tools each one called, how many tokens it used, how long it ran, and what it cost. No SDK, no proxy, no config — it just works. Anthropic closed community feature requests for this (#10164, #10388), so BurnRate is currently the only way to get per-subagent cost breakdowns.

Which AI tools are supported?

Seven providers: Claude Code, Cursor IDE, OpenAI Codex, GitHub Copilot, Windsurf, Cline, and Aider. The plugin architecture makes it easy to add new providers as they emerge.

How does licensing work?

Free tier works without any account. Pro and Team plans use a license key that you enter once. The key validates online every 24 hours with a 7-day offline grace period, so it works on planes and in secure environments.

What does the optimization engine do?

It analyzes your usage patterns across 46 dimensions: output-heavy sessions, context accumulation, session depth, operation costs, duplicate work, cache efficiency, provider arbitrage, budget overruns, and more. Each recommendation includes plain-English explanations, real-world analogies, and specific savings estimates based on your actual usage.

Can I use this at work?

Many teams can use BurnRate because raw code and session content stay local, but BurnRate is not currently SOC 2 certified. Confirm your organization's requirements before deployment. Team plans add shared aggregate dashboards.

What if I cancel Pro?

You keep the Free tier features forever. Your local data is never affected. You just lose access to the advanced analytics tabs and optimization engine.

How does team analytics work?

Each developer installs the BurnRate binary on their machine. It analyzes their local AI session data and pushes only aggregated cost metrics (no code, no session content) to your team dashboard. Admins see per-developer and per-project spending without ever accessing raw data.

Is this SOC 2 compliant?

BurnRate is not currently SOC 2 certified. Raw code and session data are processed locally. Cloud services handle authentication and billing, plus aggregate metrics when team sync is enabled.

Do you offer volume discounts?

Yes. Teams of 50+ seats qualify for custom Enterprise pricing with volume discounts, dedicated support, and custom integrations. Contact matt@getburnrate.io.