62 models · 17 providers

Your AI bill broken down
by project.

One dashboard for every LLM you use. See what each project, workflow, and feature actually costs — whether you pay per token or track subscription usage.

getpromptcost.com — dashboardlive

Total spend

$284.40

30-day spend

$91.20

Total tokens

18.7M

Sessions

342

api-backend/
gpt-4o$124.80
my-saas-app/
claude-sonnet-4-6$91.20
data-pipeline/
gemini-2-flash$48.60
internal-tools/
deepseek-r1$19.80

Who it's for

Built for everyone paying for AI

Whether you get a monthly invoice or use a subscription — if AI is part of your workflow, you need visibility.

API developers

You pay API bills

You get a $300 invoice from Anthropic. You have 8 projects. PromptCost tells you exactly which ones spent what — so you can cut the waste.

Subscription users

You use subscriptions

On Claude Code, Cursor, or ChatGPT Plus? Track session counts, token usage trends, and see which projects consume the most context over time.

Product builders

You build AI products

Running multiple workflows in production? Log every LLM call in one line of code. See cost per feature, per customer, per day.

Providers

Every major AI provider

62 models across 17 providers. One place to see it all.

Anthropic

Claude 3.5, Opus, Haiku

OpenAI

GPT-4o, o1, o3-mini

Google

Gemini 2.0, Flash, Pro

DeepSeek

R1, V3, Coder

Groq

Llama 3.3, Mixtral

Mistral

Large, Codestral

+ 11 more providers

Cohere, Reka, AI21…

Claude Code

Install once.
Tracks forever.

Add one hook to your Claude Code settings. Every session auto-logs when you stop — project name, token count, cost. Zero work after setup.

See full setup guide
~/.claude/settings.json
{
  "hooks": {
    "Stop": [{
      "hooks": [{
        "type": "command",
        "command": "bash ~/.claude/promptcost-hook.sh"
      }]
    }]
  }
}
Auto-logs on every session end

Using Cursor, Aider, or the raw API?

One API call after any LLM response. Works in any language, any framework.

await fetch("https://getpromptcost.com/api/ingest", {
  method: "POST",
  headers: { "Authorization": "Bearer YOUR_API_KEY" },
  body: JSON.stringify({
    workflow: "my-project",
    model: "gpt-4o",
    input_tokens: response.usage.input_tokens,
    output_tokens: response.usage.output_tokens,
  })
})

Everything you need

No fluff. Just the data you actually use.

Per-project breakdown

Every project tracked separately. See which codebase is burning your budget.

30-day cost trends

Daily spend over time. Know exactly when costs spiked and why.

Zero-config hook

Install once for Claude Code. Sessions log automatically when you stop.

Budget alerts

Set a monthly limit. Get notified before you blow the budget.

CSV export

Export any date range for accounting, billing clients, or analysis.

62 models supported

Every major provider. Pricing always up to date.

Stop guessing what AI costs you.

Free plan — 5 projects, 50 sessions. No credit card.

Claude Code users are up and running in under 2 minutes.