llm-usage

Local gateway for Claude Code, Codex and other AI clients on macOS, Windows and Linux: switch upstreams without touching clients, keep API keys from relays, cut off malicious tool calls, trace every request.

The one dashboard you’ve been looking for — track spend and usage across Claude, Cursor, OpenRouter, Copilot, Gemini, Codex, and more.

Real-time usage, cost & quota tracking across 50+ AI CLI tools (Claude Code, Codex, Gemini, Cursor, Copilot…) — plus remote-control your Mac's AI coding sessions from your phone. macOS menu bar · iPhone · iPad · Apple Watch · Android.

Comprehensive cost & token usage analyzer for Claude Code with detailed breakdowns by model and token type. Uses ccusage CLI tool and LiteLLM pricing data. The raw token usage data is available in ${HOME}/.claude/projects/<project-name>/<conversation-id>.jsonl files.

A Desktop Widget for Linux / Windows / MacOS that shows you your current LLM usage levels

A local-first AI workstation for managing AI tools, agents, machines, workspaces, terminals, tasks, workflows, automations, and usage. Cross-platform desktop and mobile apps with remote access, MCP, and unified AI usage analytics.

A self-hosted hosted dashboard that visualizes your opencode CLI usage analytics

Read-only Android companion for monitoring AI token usage, provider limits, and trends from Token Monitor, with home-screen widgets over your private network.

Claude Code plugin for Claude Desktop: a band above the prompt with this chat's share, the 5-hour window and the week, run-out estimates; a usage line under every answer; a panel with top chats. By Roman Chernogorov & Danil Martsinkovskii.

本机 AI 编码 Agent 的 Token 用量总览 · 零依赖 · 支持 Proma / Pi / oh-my-Pi / Claude Code / Claude Desktop / Codex / ZCode / DeepSeek Harness / OpenCode

Token Pie runs locally on each developer's machine, reads telemetry Copilot already writes, Nothing leaves the machine. No server, no account, no telemetry of our own.

Gateway engine for Claude Code, Codex and other AI clients, inside ThinkWatch Lite or on a Linux server: credential redaction, tool-call inspection, rule-based routing, per-request cost. Rust, MIT.

Claude, Codex & Gemini usage limits in your macOS menu bar. Native Liquid Glass. Companion to lord-claude.

Local-first LLM subscription usage for Hyprland, OpenAI / GLM / Grok / Claude in your bar (CLI + Quickshell)

Local LLM token usage analytics CLI with a dashboard and SQL-free custom reports. Supports Claude, Codex, OpenCode, ZCode, WorkBuddy, Zhipu-AutoClaw, MiMo Code/Desktop, and DeepSeek Harness (DSH). | 本地 LLM 用量分析：可视化仪表盘、无需 SQL 的自定义多维报表，支持 DSH 等多种 AI 客户端。

A fun skill that turns your coding agents' local history into a shareable card with your tokens, their API list price, your plan limit usage, and a roast.

DeepSeek Harness Token 用量统计面板：模型/渠道聚合、余额配额、精确归档与费用估算

DeepSeek Harness(dsh)持久化 Token 计费插件:按 token 计量每次模型调用的成本,人民币默认价,账单落盘跨会话累计,含统计栏金额、设置面板与 billing_report 工具。

Rust logistics exception handling with an account-enforced model spend cap.

Read the local WorkBuddy ledger and render it into a five-window report (today / yesterday / 7 days / 30 days / all) with token, credit and cost breakdown by session / model / workspace. Read-only, offline, zero LLM calls.