Tokens for GoodGuides
Everything we have written about getting more out of a Claude subscription, from usage limits and idle capacity to MCP servers and ways to put spare compute toward something useful.
Claude plans and usage limits
- Is Claude Max worth it? A straight answer on whether Claude Max is worth it: who it actually pays off for, the usage math, and what to do with capacity you pay for but never use.
- Claude Max vs Pro: which is worth it? An honest Claude Max vs Pro comparison by usage pattern. See which plan fits your workload and why most subscribers leave idle headroom either way.
- Claude Max usage limits, explained A plain-English reference to how Claude Max usage limits and resets work, how to genuinely use them, and what to do with the surplus you never drain.
- Claude weekly usage limits, explained How Claude weekly usage limits actually work: the rolling 5-hour window, the two weekly caps, when they reset, and what happens to headroom you never use.
- What the approaching weekly limit warning means in Claude What the approaching weekly limit warning in Claude actually means, how much room you really have left, how to keep working, and when your cap resets.
- Do Claude tokens roll over? Claude tokens do not roll over; usage limits reset each cycle and unused headroom expires. How the resets work and what to do with capacity you would lose.
- Are you wasting your Claude subscription? Most Claude subscribers never touch their weekly ceiling. Four quick signs you over-bought, when downgrading makes sense, and how to make idle capacity useful.
- Should I cancel Claude Max? Thinking about whether to cancel Claude Max? A fair guide to the signs you should downgrade, when it is worth keeping, and how to make idle capacity count.
- What to do with unused Claude capacity Your Claude plan resets its usage limits and does not roll over, so unused capacity is gone. Here are productive ways to use the headroom you never drain.
- What is the Claude 5-hour session limit? What the Claude 5 hour session limit is, what starts the window, what happens when you hit it, when it resets, and how it differs from the weekly caps.
- Does Claude have a daily usage limit? Claude has no daily usage limit. Here is what people are really asking, how the five-hour window and weekly caps work instead, and when each one resets.
- When do Claude limits reset? When Claude limits reset: weekly caps at a fixed time assigned to your account, the five-hour window from your first message, both shown in your local time.
- How to check your Claude usage Where to look: Settings > Usage in the Claude app and /usage in Claude Code, what the percentages mean, and what "increase my weekly limit" really resolves to.
- Claude reached its tool-use limit for this turn: what it means What "Claude reached its tool-use limit for this turn" means: a per-turn cap on tool calls, not your usage limit. Why it fires and how to restructure the ask.
- Claude usage limit reached: which limit did you hit? What "usage limit reached" means in Claude, how to tell which clock you hit, how long the wait is, what to do right now, and why "at capacity" is something else.
- Claude Pro usage limits, explained How Claude Pro usage limits work: the five-hour session window, the weekly caps, Pro versus Free and Max, Claude Code on Pro, and how to read your own numbers.
- The Claude Opus usage limit Opus has its own weekly ceiling on top of the all-models cap. What "Opus limit reached" means, why switching to Sonnet works instantly, and how to plan Opus work.
- Claude Max 5x vs 20x: which tier do you need? Claude Max 5x vs 20x compared: what each tier actually changes, who needs the top tier, a self-check by weekly pattern, and the cheaper middle path.
- Claude Code usage limits, explained How Claude Code usage limits work on a subscription: the same five-hour window and weekly caps as chat, why agentic work spends them faster, and how to plan runs.
Claude cost and API pricing
- Claude Pro vs API: the real cost A clear Claude Pro vs API cost comparison for developers. See when a flat subscription beats metered tokens, and what the surplus on light days is worth.
- Claude API pricing, explained What the Claude API actually costs: per-model input and output rates as of August 2026, caching and batch discounts, and three worked monthly examples.
- Does Claude Pro include API credits? No. Claude Pro and the Claude API are separate products on separate billing. What each one covers, where Claude Code blurs the line, and which one to pick.
- Claude Code: subscription or API billing? Claude Code runs on a Claude subscription login or on Console API billing. What each path changes, how Claude Code picks one, when to switch, and how to run both.
- Claude rate limit error: which limit did you actually hit? A Claude rate limit error means one of two different things: an API 429 rate_limit_error, or a subscription usage limit. How to tell which you hit, and how to fix each.
Idle capacity and automation
- When is the best time to use Claude? Does Claude have busy times? How peak hours, the five-hour window, and weekly caps really interact, and how to schedule heavy work for quiet hours.
- Productive uses for idle Claude Code Productive uses for idle Claude Code: turn spare subscription capacity into tests, docs, research, and vetted nonprofit work instead of letting it go to waste.
- Best Claude Code overnight tasks The best Claude Code overnight tasks to queue before bed: test coverage, dependency audits, research briefs, refactors, and one that does real good while you sleep.
- Claude Code scheduled task ideas Practical Claude Code scheduled task ideas: recurring audits, digests, cleanups, and a standing contribution that vets nonprofits while you sleep.
- Claude Code cron job ideas Claude Code cron job ideas: recurring agent tasks worth a fixed schedule, from nightly audits to digests, plus a clean give-back cadence example.
- How to schedule Claude Code How to schedule Claude Code: recurring cloud routines with /schedule, local cron with headless mode, prompt-writing tips, and a set-and-forget worked example.
- What to automate with Claude Code What to automate with Claude Code: the categories of work worth handing off, from tests and docs to research, plus the useful-while-you-sleep pick.
- Set-and-forget Claude Code agents Set-and-forget Claude Code agents that need no babysitting: hands-off jobs with clear stopping points, including one that does real good while it runs.
- Claude Code subagent ideas A practical list of Claude Code subagent ideas: research sweeps, doc reading, log analysis, test runs, and a scheduled Claude Code subagent that does good.
- Claude peak hours: what the policy actually is now Does Claude have peak hours? As of August 2026 there is no peak-hours limit reduction for Pro or Max. What changed, what still varies by hour, and how timezones work.
- How long can Claude Code run? Claude Code has no hard time limit; a session can run for hours. What stops a long run is the five-hour usage window, weekly caps, and work that needs a human.
- How to reduce Claude Code token usage Where tokens actually go in a Claude Code session, and the habits that cut waste: scoping, /clear and /compact, a lean CLAUDE.md, model choice, and small tool output.
MCP servers
- Best MCP servers for good (2026) A curated roundup of MCP servers that do social good: public-interest data, charity due-diligence tooling, and one that turns spare AI capacity into nonprofit research.
- Best MCP servers in 2026 The best MCP servers to install in 2026: practical essentials like Filesystem, GitHub, Context7, and Playwright, plus the standout philanthropy pick for giving back.
- Useful Claude Code MCP servers The most useful MCP servers for Claude Code: practical everyday picks like Filesystem, GitHub, Context7, and Playwright, plus a give-back server for spare capacity.
- Free MCP servers worth installing A roundup of free MCP servers worth installing: Filesystem, GitHub, ProPublica 990 lookups, the Candid connector, and a free give-back server for spare AI capacity.
- Charitable MCP servers Charitable MCP servers for 2026: verify charities, read 990 financials, search funder data, and turn spare AI capacity into nonprofit research.
- MCP servers for nonprofits A practical directory of MCP servers for nonprofit research: verify EINs, pull 990 financials, search funder data, and contribute spare AI capacity to vet charities.
- How to add an MCP server to Claude Code A practical guide to adding MCP servers to Claude Code: the claude mcp add command, local vs remote transports, scopes, and a worked example you can copy.
- MCP servers for accounting and taxes An honest guide to MCP servers for accounting and tax work: which platforms ship official servers, which categories are thin, and the compliance limits that matter.
AI for good and giving
- AI for good projects worth knowing in 2026 A roundup of real AI for good projects in 2026, from disaster mapping to nonprofit research, plus a low-friction way developers can join in.
- How to use AI for good How to use AI for good with concrete actions for individuals and developers, from skilled volunteering to giving spare AI capacity in one command.
- AI for nonprofits: practical uses and how to help How nonprofits are actually using AI in 2026: writing, research, and operations, the honest caveats, and how developers can contribute spare AI capacity to help.
- What is compute philanthropy? Compute philanthropy is giving spare computing power to good causes. From Folding@home to contributing idle AI capacity, here is what it means and how to start.
- Volunteer computing for AI Volunteer computing for AI explained; how the Folding@home and BOINC model maps to spare AI capacity for nonprofit research through Tokens for Good.
- How to contribute spare compute to good causes How to contribute spare compute to good causes; from classic volunteer computing to giving unused AI capacity for nonprofit research at no extra cost.
- How to contribute your spare AI capacity A step-by-step guide to putting your unused Claude tokens to work vetting nonprofits with Tokens for Good. No extra cost, not crypto, no coins.
- How to give to charity without money How to give to charity without money; contribute time, skills, blood, goods, and spare compute, with a developer-friendly way to give AI capacity.
- Ways developers can give back Practical ways developers can give back; open source, mentoring, skilled volunteering, and giving spare AI capacity to nonprofit research at no extra cost.
- Earning to give for developers Earning to give for developers explained; how the effective-altruism idea works, its tradeoffs, and how giving spare compute complements it at no cost.
- How to vet a charity before you give Learn how to vet a charity the right way: check financials, Form 990, transparency, and real outcomes, and use free tools like GuideStar and ProPublica.
About Tokens for Good
- What is Tokens for Good? Tokens for Good lets developers contribute spare Claude capacity to research and vet nonprofits. Here is what it is, how it works, and why it is not crypto.
- How Tokens for Good works A clear walkthrough of how Tokens for Good turns spare AI capacity into vetted nonprofit reports: research twice, validate, consolidate, score, human review.
- Is Tokens for Good legit? A straight answer on whether Tokens for Good is safe and legitimate: what it accesses, that it uses your own revocable key, and how its output is reviewed.
- Tokens for Good review (2026) An honest Tokens for Good review: how it works, what is good, who it is not for, and what to know before you contribute spare Claude capacity to vet nonprofits.
- Tokens for Good vs Folding@home How Tokens for Good adapts the Folding@home volunteer-computing model for the AI era: idle AI capacity vetting nonprofits instead of idle GPUs folding proteins.
Put idle Claude capacity to work
Tokens for Good turns the headroom you never drain into cited research on high-impact nonprofits, running on your existing subscription.
Get started