CodexBar — What is it?

CodexBar is a menu bar application that provides usage statistics for various AI coding providers, enhancing productivity and cost management for developers.

⭐ 17,819 Stars 🍴 1,459 Forks Swift MIT Author: steipete
Source: Official README and linked documentation View on GitHub →

Why it matters

CodexBar is being discussed because its support for multiple AI coding providers, its privacy-first approach by reusing existing sessions, and its cross-platform availability on macOS and Linux. The project's unique technical choice of a Qt desktop app for Linux and its integration with various AI services differs.

Source: Official README and linked documentation

Core Features

Provider Usage Stats

CodexBar offers detailed usage statistics for various AI coding providers, including credits, spend, and cost scans, allowing developers to monitor their usage and costs effectively.

Source: Official README and linked documentation
Live Status Updates

The application provides live status updates for AI providers, including incident badges and indicator overlays, ensuring developers are aware of any service disruptions.

Source: Official README and linked documentation
Privacy-Focused Design

CodexBar is designed with privacy in mind, reusing existing provider sessions without storing passwords, which appeals to users concerned about data security.

Source: Official README and linked documentation

Architecture

The architecture of CodexBar suggests a modular design with separate modules for different AI providers. It uses a Qt framework for the desktop app, with a clear separation of concerns in the code structure. Data flow is managed through a central configuration system, and key technical decisions include the use of OAuth for authentication and a CLI for configuration.

Source: Official repository tree and dependency files

Project Knowledge Graph

Knowledge graph: project (center) + core features (inner hexagons) + key dependencies (outer chips) OAuth API keys CLI Provider Usage Stats Live Status Updates Privacy-Focused DesignPrivacy-Focused Des… CodexBar Project Core feature Key dependency

Tech Stack

LanguageSwiftFrameworkQt 6
OAuthAPI keysCLI
macOS, Linux desktop
Source: Official README and linked documentation

Quick Start

macOS: Download from GitHub Releases or install with Homebrew. Linux: Download release binaries from GitHub Releases, install runtime dependencies, and run the installer script. CLI Tarballs: Install using Homebrew or download and extract release tarballs.
Source: Official README and linked documentation

Use Cases

CodexBar is suitable for developers who use multiple AI coding providers and need to monitor their usage and costs. It is useful in scenarios where developers need to plan around resets, manage credits and spending, and ensure privacy in their usage of AI services.

Source: Official README and linked documentation

Strengths & Limitations

Strengths

  • support for multiple AI coding providers
  • Privacy-focused design
  • Cross-platform availability

Limitations

  • Limited documentation for some features
  • May require technical knowledge to set up and configure
Source: Official README and linked documentation

Latest Release

The official Sparkle appcast with changelog display v0.64.1 was released on 2026-09-22, fixing provider, Linux, Kimi and menu bar display issues. main's version.env is already 0.64.2 and changelog is marked as Unreleased, so this page does not consider the change of 0.64.2 as a delivered feature.

Source: github.com/steipete/CodexBar/blob/main/appcast.xml;https://github.com…

Verification and evidence

status: docs-only  |  checked_at: 2026-09-23  |  environment: Read the official README, v0.64.1 appcast/CHANGELOG, Providers, Architecture, Configuration, Keychain, CLI, Linux installation, Package.swift, and license; shallow clone of the official main of 2026-09-22. No app or CLI installed.  |  limitations: No local installation or runtime verification was performed. Signatures, notarization, Homebrew, provider login, Keychain, cookie decryption, API requests, usage and cost accuracy, the updater, Linux desktop, CLI, plugins, and local history performance were not tested.  |  observed_output: The official appcast points to the release of the asset v0.64.1, changelog date is 2026-09-22; the main version.env has been 0.64.2 Unreleased. README currently lists 78 providers and declares that browser cookies are opt-in and that provider tokens have restricted permission profiles.

Source: github.com/steipete/CodexBar;https://github.com/steipete/CodexBar/blo…

Comparison

Compared with AgentsView, CodexBar focuses on the display of credits, resets, balances, and local costs across providers; AgentsView focuses on long-term archiving of full sessions, full-text searches, APIs, and MCPs for multiple coding agents. CodexBar is more direct when you just want to look at balance and billing trends; AgentsView is more appropriate when searching for historical conversations and project records. Both centrally read sensitive local data, and data sources should be restricted separately.

Source: github.com/steipete/CodexBar;https://github.com/kenn-io/agentsview

Verdict

CodexBar is a tool for developers looking to manage their usage of AI coding providers efficiently and securely. Its feature set and privacy focus make it a strong choice for those who require detailed usage statistics and cost management across multiple platforms.

Frequently Asked Questions

What did the documentation review verify for CodexBar on 2026-09-23?

The official appcast points to the release of the asset v0.64.1, changelog date is 2026-09-22; the main version.env has been 0.64.2 Unreleased. README currently lists 78 providers and declares that browser cookies are opt-in and that provider tokens have restricted permission profiles.

What evidence boundary applies to this CodexBar review?

The review environment was: Read the official README, v0.64.1 appcast/CHANGELOG, Providers, Architecture, Configuration, Keychain, CLI, Linux installation, Package.swift, and license; shallow clone of the official main of 2026-09-22. No app or CLI installed.

Sources and review scope
This review is based on the project's public documentation. Versions, installation commands, and feature boundaries can change upstream; check the cited sources and current release before use. Sources checked on: 2026-09-23 13:55.

Reference materials: README, GitHub API, dependency files