NEW Spend targets — plan and track by group

Your bill went up. Find out why. Stop it happening again.

Every LLM call tagged by group, user, and model. Set spend limits by tier. Plan where your AI budget goes — and know when it drifts.

Overview · Last 30 days
Live
Total spend
$2,847
↓ 12% vs last mo
Blocked calls
143
budget limits enforced

search costs 5× more than chat per call — 18k vs 3.7k avg tokens.

Trusted by teams
shipping AI in prod
5 minmedian time to first insight
$0infrastructure to manage
1 lineto start tracking
0mslatency added to your calls
Architecture

No proxy.
No middleman.
No new failure mode.

Most LLM observability tools sit between your app and your provider. Every call routes through their servers — adding latency, a new dependency, and a question your security team will eventually ask. We work differently.

Proxy-based tools

Your requests travel through a third-party server before reaching the provider.

Your app
 
Their proxy
third-party infra
Provider
 

Latency added — every call takes an extra hop through their infrastructure

New failure mode — their outage becomes your production incident

Data exposure risk — your requests physically pass through their servers

LLM Cost Tracker

Your requests go directly to the provider. Always. The SDK reads metadata after the call resolves.

Your app
 
Provider
direct
SDK wrapper
reads response

Zero latency impact — the SDK reads token counts after your call resolves

No new dependency — if our service is down, your LLM calls are completely unaffected

Prompts never leave your stack — we log cost metadata only, not your conversations

Direct to provider — your API calls never touch our servers

Prompt content never logged — token counts and metadata only

Our downtime is never your downtime — zero coupling to your call path

vs provider dashboards

Anthropic tells you when your bill is high. We tell you why.

Provider dashboards show you what you owe. LLM Cost Tracker shows you why — and what to do about it.

Provider dashboard
LLM Cost Tracker
Cost by feature / group
Cost by user
Cost by model
Per-call attribution
Real-time data
24–48h lag
✓ live
Spend targets per group
Budget alerts
Account-level only
✓ per group, per user
SDK-level enforcement
✓ block / warn / dry run
No proxy / zero latency
N/A
✓ always direct

Anthropic Console, OpenAI Platform

Observe · Plan · Enforce

Your AI invoice is one line item. Your dashboard shouldn't be.

Three capabilities. One SDK. No infrastructure to manage.

01
Observe
Every LLM call broken down by group, user, model, and prompt version. See exactly which part of your product is driving your bill — in real time.
Learn more →
02
Plan
Set a monthly dollar target per group. Track actual vs plan. Get alerted when spend is projecting over before the month ends.
Learn more →
03
Enforce
Set hard spend limits per user and per tier. The SDK enforces them before the API call is made — block, warn, or dry-run. No infrastructure required.
Learn more →
From the field

What teams find when they turn it on.

"

We noticed our Anthropic bill going up and had no idea which searches were expensive, which were cheap, or why. We dropped in LLM Cost Tracker to find out.

Within 10 minutes we spotted a 5× cost variance between two searches — $0.0123 vs $0.0608 — driven by token count differences we couldn't see before.

B

Brian

Founder, contractclues.com

Privacy

We track your costs. Not your conversations.

The SDK reads token counts and metadata from the response object after your call resolves. It never sees your prompts, your users' messages, or your model outputs.

CapturedWhat the SDK logs
Token counts
input_tokens, output_tokens — from the API response object
Model name & calculated cost
e.g. claude-sonnet-4-6 · $0.0061 — derived from token counts
Latency
Wall-clock ms from call start to response complete
Your attribution tags
group, userId, tier, promptVersion — values you pass in explicitly
Never touchedWhat stays in your app
Prompt content
System prompts, user messages — never read, never logged
Model outputs
Completions, streamed text, tool call results — not captured
Your API keys
The SDK uses your existing client — we never see your credentials
PII or user data
No names, emails, or identifying content — only the IDs you pass
Your data stays in your stack
Events log directly to your own Supabase or Postgres database. Nothing hits our servers unless you opt into cloud hosting.
Fully self-hostable
Run the entire stack in your own infrastructure. No data residency concerns. Required for fintech, healthcare, or any regulated environment.
Open source SDK
Read every line before installing. github.com/llmcosttracker/sdk →
Pricing

Simple pricing. No per-seat games.

Start free. Upgrade when your usage grows.

Free
$0
For solo devs evaluating the tool.
  • 10,000 events / mo
  • 1 project
  • Full dashboard access
  • Call log + group breakdown
  • Spend targets
  • API key
Start free
Growth
$29/mo
For products with real LLM volume.
  • 500,000 events / mo
  • 5 projects
  • Everything in Starter
  • Spend enforcement + tier templates
  • Prompt version tracking
  • Slack alerts
  • Priority support
Start free
Enterprise
Custom
Custom volume, private deployment.
  • Unlimited events
  • Unlimited projects
  • Self-hosted option
  • SSO + SAML
  • SLA · dedicated support
Contact us

Need the full breakdown? See all features by plan →

© 2026 LLM COST TRACKERhello@llmcosttracker.com