Free · open source · no API keys

LLM cost tools for developers

Current prices for 14 OpenAI, Anthropic and Google models, inside your AI assistant, your pull requests and your own code. MIT-licensed code, CC BY 4.0 data, checked weekly against each provider's pricing page.

In short

Free, open-source LLM cost tools from AiGanak for developers. The AiGanak LLM Cost MCP server (npm @aiganak/llm-cost-mcp, MCP Registry io.github.AiGanak/llm-cost) gives Claude, Cursor and other AI assistants current OpenAI, Anthropic and Google prices, cost estimates, model comparisons and the language token tax. The LLM Cost Check GitHub Action (AiGanak/llm-cost-action) finds model IDs in code and flags price rises, legacy models and cheaper options on pull requests. Open price data in JSON and CSV with a Python package on GitHub (AiGanak/llm-prices). No API keys; MIT code, CC BY 4.0 data.

MCP server →GitHub Action Open data

MCP server · npm @aiganak/llm-cost-mcp

Ask your AI assistant what a workload will cost

Works with Claude Desktop, Claude Code, Cursor, VS Code, Windsurf and any other MCP client. Ask "Which model is cheapest for a Tamil support bot?" or "What will 50,000 requests cost on Claude Sonnet 5.5 vs GPT-6.1 Sol?" and get an answer priced from current, sourced data.

npm · Source on GitHub · Listed in the official MCP Registry as io.github.AiGanak/llm-cost

Claude Desktop, Cursor and others

Add to claude_desktop_config.json or ~/.cursor/mcp.json:

{
  "mcpServers": {
    "aiganak-llm-cost": { "command": "npx", "args": ["-y", "@aiganak/llm-cost-mcp"] }
  }
}

Claude Code

claude mcp add aiganak-llm-cost -- npx -y @aiganak/llm-cost-mcp

Tool	What it answers
`list_models`	Current models with standard, cached-input and batch prices per million tokens
`get_model_price`	Full pricing for one model: cache writes, long-context rates, announced price changes, source page
`estimate_cost`	Cost of N requests on one model, with caching, batch and future-dated prices
`compare_models`	The same workload priced on every model, cheapest first
`token_tax`	How many more tokens a language takes than English, per model family (22 languages)
`estimate_text_cost`	Tokens and cost for a piece of text in any measured language, script detected automatically
`price_changes`	Dated, sourced log of LLM API price changes

GitHub Action · AiGanak/llm-cost-action

See LLM price changes in every pull request

LLM Cost Check finds the OpenAI, Anthropic and Google model IDs in your code and comments on pull requests with:

each model's price per million input and output tokens
announced price rises, with the date
legacy and preview models
a cheaper model in the same tier
the price change when a pull request switches models, in dollars a month if you give your volume
model IDs that are not in the current price list, such as older generations

GitHub Marketplace · Source and docs on GitHub

Add one file

# .github/workflows/llm-cost-check.yml
name: LLM Cost Check
on:
  pull_request:
  schedule:
    - cron: "0 6 * * 1"   # weekly, catches newly announced price changes
permissions:
  contents: read
  pull-requests: write
jobs:
  llm-cost:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v5
      - uses: AiGanak/llm-cost-action@v1.0.0
        with:
          fail-on: none   # or: price-increase, legacy, unknown, any

Open data · AiGanak/llm-prices

Use the prices in your own code

JSON and CSV for every current model: standard, cached-input and batch prices, cache writes, long-context rates and announced price changes, each with its source page. Plus a dated log of price changes and the measured language token tax for 22 languages. A Python package and command-line tool are included.

Data and docs on GitHub · Browse the prices · Language token tax

Fetch or install

curl -s https://cdn.jsdelivr.net/gh/AiGanak/llm-prices@main/data/prices.json

pip install "git+https://github.com/AiGanak/llm-prices"

FAQ

Developer tools questions

Can't find it? Press ⌘K and ask.

Do I need an API key?

No. The MCP server, the GitHub Action and the data use public list prices, so none of them needs an OpenAI, Anthropic or Google key or an AiGanak account.

Where do the prices come from and how fresh are they?

From each provider's own pricing page, checked weekly. Every model entry links to its source page, and announced price changes are listed with their dates. The MCP server and the GitHub Action fetch the latest data each time they run.

Can I use the data in my own product?

Yes. The code is MIT-licensed and the data is CC BY 4.0: use it commercially, with a credit to AiGanak.

Does the GitHub Action read my code or send it anywhere?

It runs inside your own GitHub Actions job, scans your files there and only downloads the public price list. Nothing from your repository is sent to AiGanak.

Want the whole bill analysed?

The LLM Cost Report reads your real usage, prices every current model and gives you a ranked savings plan with code, in two working days.

See the LLM Cost Report →Free cost check