Coding
Track LLM API spending per agent/session with budget alerts and CSV export
Track LLM API spending per agent/session with budget alerts and CSV export
Estimate and tally LLM API spending per agent/session with budget alerts and CSV export.
Keywords: cost, tracking, llm, budget, cli, finance, python, open-source, agent, automation, MIT
Part of the itsPremkumar Hermes / OpenClaw / Paperclip agent stack — 31 free, MIT-licensed, CI-tested agent-native tools.
Token spend balloons silently across agents and models with no per-agent accountability. LLM Cost & Token Tracker solves this: Estimate and tally LLM API spending per agent/session with budget alerts and CSV export.
Best for: Founders, finops, and agent operators running multi-agent LLM workloads on GPT/Claude/Gemini.
# Requires Python 3.8+. No pip install needed.
curl -O https://raw.githubusercontent.com/itsPremkumar/agent-cost-tracker/main/agent_cost_tracker.py
# Or copy the file anywhere — it's self-contained.
python agent_cost_tracker.py self-test # prove it works end-to-end
python agent_cost_tracker.py tally --help # tally subcommand
python agent_cost_tracker.py estimate --help # estimate subcommand
| Alternative | Why this skill is better |
|---|---|
| Cloud billing dashboards | Per-agent, per-session granularity you control locally. |
| Mental math on token counts | Price tables + estimation remove the guesswork. |
| Spreadsheet tracking | CSV export drops straight into your existing finance stack. |
Q: Which models are supported?
A: gpt-* (OpenAI), claude-* (Anthropic), gemini-* (Google) with built-in price tables.
Q: Where does it get tokens?
A: From agent run logs (tally) or from --prompt/--completion estimates.
Q: Does it call the API?
A: No. It is an estimator/tally from logs and inputs; no network calls.
Q: Can I export to CSV?
A: Yes — CSV export is built in for spreadsheets and BI tools.
Built and maintained by @itsPremkumar (Chennai, India · serving developers worldwide). Free for individuals and teams everywhere. Documentation in English; tool output is locale-neutral.
# .github/workflows/verify.yml
name: Verify
on: [push]
jobs:
verify:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Self-test agent-cost-tracker
run: python agent_cost_tracker.py self-test
Free + MIT-0 (free, modifiable, no attribution required). Sponsor if useful:
⭐ Star on GitHub
Track every token and dollar your OpenClaw agent spends, set budget walls that actually refuse calls, and replay runs on free local models. Local-first: all data stays on this machine. Use when the user asks about cost, spend, tokens, budgets, usage reports, or what the agent has been doing.
Record every tool call into agent-tracer (self-hosted FastAPI), then run regression testing (golden case → drift detection), attribute token costs by tool/model/agent, and surface recurring error root causes across sessions. Use when an agent must log its actions, verify behavior hasn't drifted, report spend, or debug repeated failures.
Audit and reduce AI agent runtime spend in dollars. Use for AI costs, agent spend, token waste, runtime attribution, detector coverage, and FinOps. Works with OpenClaw, Hermes, QM, Claude Code, Cursor, and generic event ingest.
Build or modernize TypeScript CLIs for AI agents with @renxqoo/agent-cli-sdk. Use when a user wants a new command-line tool, an API or internal service wrapped as a CLI, or an existing agent-cli-sdk app extended with authentication, structured output, typed errors, pagination, pipes, Skill distribut
Tiny dependency-free TTL cache that skips a repeat LLM/API call entirely when the same input recurs. Fixes a common cost leak in agents that re-score/re-anal...
LLM gateway & AI completion API for agents — call frontier models (Claude, GPT, DeepSeek) across cheap/pro/ultra tiers from one endpoint, pay-per-call in USD...