Guides

Run Claude Code lean.

Straight, factual guides on how the usage limits work, where your tokens go, and how to make the plan you already pay for last the whole week. The free audit CLI is the measurement starting point.

• Usage limits Claude Code usage limits: 5-hour window, weekly cap, and resets How Claude Code usage limits work — the rolling 5-hour window, the weekly cap, per-plan differences, and when each resets — plus how to see your burn before you hit the wall.
• Alternatives A Headroom alternative for Windows, macOS, and Linux Headroom is a genuinely good macOS tool for cutting Claude Code token usage. Here's an honest look at how Hypermile differs — including Windows, macOS, Linux, and WSL2.
• Token waste Claude Code costs: what you actually pay and how to cut it What Claude Code actually costs on a subscription versus the API, where the tokens really go — tool output, file re-reads, and sub-agents — and how to stretch the plan you already pay for on Windows, WSL, macOS, or Linux.
• WSL Tracking Claude Code token usage in WSL2 WSL2 changes where your Claude Code transcripts live and how a local proxy has to reach them. Here's what's different about token tracking inside WSL — and how Hypermile is built to support it natively.
• Comparison Hypermile vs Headroom: Which is better? (2026) A factual, dated feature-by-feature comparison of Hypermile and Headroom — platforms, install size, forecasting, transparency, free tier, and pricing — with honest notes on where each one leads.
• Usage limits Codex usage: how to check it, what counts, and when it resets How to check your Codex usage from the CLI and your ChatGPT account, what counts toward the 5-hour window and weekly cap, when each resets, and how to stretch the allowance you already pay for.
• Comparison Hypermile vs RTK: Which is better? (2026) A factual comparison of Hypermile and RTK — CLI output compression vs a usage copilot with forecast, inspector, and multi-OS desktop — with honest notes and no fabricated benchmarks.
• Usage limits Codex weekly limit: when it resets and what happens when you hit it The Codex weekly limit is a separate 7-day cap on top of the 5-hour window. When it resets, what happens when you hit it, and how to make the week last.
• Windows Claude Code on Windows: the complete setup and troubleshooting guide A complete guide to running Claude Code on Windows — install paths, PowerShell vs Git Bash gotchas, CRLF and path issues, where settings and transcripts live, and how to keep your usage under control.
• WSL Claude Code WSL2: setup and performance guide Install and run Claude Code on WSL2 the fast way — distro setup, why /mnt/c is slow, and how keeping the project in your Linux home directory fixes performance.
• Comparison Hypermile vs ccusage: a fair comparison ccusage is an excellent free CLI for reporting Claude Code token usage. Here's an honest, dated comparison with Hypermile — what each does well, where they overlap, and which fits your workflow.
• Roundup Claude usage trackers compared (2026) Every way to track your Claude and Claude Code usage — CLI tools, browser extensions, statusline options, and always-on desktop tracking — compared honestly, with a feature table and install commands.
• Usage limits Claude Max limits: how Max plan usage caps work Claude Max plan limits explained — how Max differs from Pro on the 5-hour window and weekly cap, what higher multipliers buy you, and how to stretch the allowance you already pay for.
• Usage limits Claude Code 5-hour limit: how the rolling window works What the Claude Code 5-hour limit actually is — how the rolling window starts, when it resets, how it differs from the weekly cap, and what to do when you hit it mid-session.
• API limits Anthropic rate limits: API 429s, tiers, and how they differ from Claude usage caps Anthropic rate limits explained for the API — what HTTP 429 means, how usage tiers work, and why they are not the same as Claude Code's 5-hour and weekly subscription limits.
• Usage limits Claude weekly limit: what it means and when it resets The Claude weekly limit is a fixed 7-day ceiling on top of the 5-hour window. Here’s what the message means, when it resets, and how to make the week last.
• Usage limits Claude usage limit reached: what it means and when it resets Seeing “usage limit reached” in Claude or Claude Code? Here is exactly what the message means, which clock fired (5-hour vs weekly), and when you can work again.
• Token usage Claude Code token usage: what counts and how cache reads work What counts toward Claude Code token usage — messages, tool output, file reads, sub-agents — and how cache-read weighting changes the bill relative to fresh input.
• Coming connectors Gemini CLI usage limits: what we know (and what Hypermile doesn’t support yet) Gemini CLI usage limits overview with honest pre-positioning: how Google’s agent CLI metering differs in spirit from Claude Code, and a clear note that Hypermile’s Gemini connector is not shipping yet.
• Comparison Codex vs Claude Code (2026): limits, pricing, and token burn compared A straight comparison of OpenAI Codex and Claude Code — how each meters usage, what the plans cost, which one burns tokens faster on agent work, and how to decide without guessing.
• Comparison Claude Code vs Cursor (2026): pricing, limits, and when each wins Claude Code and Cursor solve overlapping problems in different shapes — a terminal agent versus an AI-native editor. Here is how they differ on workflow, metering, and cost visibility, and why many developers run both.
• Usage limits How to check your Claude usage (Pro, Max, and API — 2026) Every place your Claude usage is visible — the web app, the Claude Code CLI, and API billing — what each surface actually tells you, and which questions none of them answer.
• Comparison OpenCode vs Claude Code (2026): open-source agent vs the incumbent OpenCode is an open-source terminal coding agent; Claude Code is the incumbent subscription product. The real trade-off is not quality — it is who controls the model choice and who absorbs the cost.
• Alternatives Claude Code alternatives (2026): what to switch to — and when not to The real alternatives to Claude Code — Codex, Cursor, open-source agents — what each trades away, the costs of switching people forget to count, and the cheaper fix if you are only leaving because of usage limits.
• Usage limits Claude Pro usage limits (2026): what you get and when it runs out How the Claude Pro plan meters you — the rolling 5-hour window, the weekly cap, why heavy Claude Code sessions exhaust Pro faster than chat does, and how to tell whether you need Max or just less waste.
• Context The Claude Code context window (2026): what fills it and why it matters What the context window actually holds during a Claude Code session, why it fills faster than you expect, what happens when it runs out, and how context pressure and usage limits are related but not the same problem.
• Context Claude Code compact (2026): what it does and what it costs you What compaction actually does to a Claude Code session, when it triggers, what you lose when it runs, and how to reduce how often you need it in the first place.