Guides for tracking AI coding usage
FeaturedStart here
Track Antigravity CLI, Pi / OMP, and Muse Code usage
Three new sources for local usage history, with setup steps and clear coverage limits.
Latest guides
Check usage, understand your limits, and find your history. Short, practical guides for all 14 providers supported by UsageAtlas.
Download UsageAtlas for freeFree and open source on Windows, macOS, and Linux. RSS feed
- Start here
Track every AI coding tool in one dashboard
One window instead of six dashboards, built from local session data.
- Claude Code
Does Claude Pro have a limit?
The direct answer, the two windows, and what the limit actually blocks.
- Claude Code
Is Claude Pro unlimited?
A direct answer, the reason the caps exist, and what changes on higher tiers.
- Cursor
When do Cursor limits reset?
One monthly reset for included usage, plus a separate on-demand cycle.
- Start here
AI coding spend tracker
Fixed seats plus variable usage, reported as two numbers rather than one.
- Codex
Codex weekly limit vs 5-hour limit
Two windows, two different clocks. What each Codex limit measures and how to plan around it.
GitHub Copilot
GitHub Copilot premium requests explained
The legacy unit, what consumed it, and how it maps to the newer credit model.
GitHub Copilot
How to check Copilot credits in VS Code
The editor shortcut, the account page it links to, and the gaps.
See more
GitHub Copilot
How to see remaining Copilot credits
Where the balance lives, what the reset date means, and why the numbers disagree.
- Claude Code
Claude usage credits explained
A separate balance from the plan allowance, and a different reset rule.
- Claude Code
Claude Pro usage: how much you actually get
Capacity is a function of model, context, and attachments, not a message quota.
- Claude Code
Claude Max 20x limits
A larger allowance, the same two windows, and a clear test for whether it pays.
- Claude Code
Claude Code token usage report
The fields, the location, and a report that is honest about estimates.
- Claude Code
Cache read tokens explained
What a cache read is, how it is priced, and the habit that increases it.
- Claude Code
Claude Code usage dashboard
The four panels that turn local logs into decisions.
- Cursor
Cursor usage dashboard
Four panels that answer whether you can keep working and what it will cost.
- Cursor
Cursor runs out after 3 messages: why
Three messages that each make twenty model calls are sixty requests.
Antigravity CLIAntigravity quota reset times
Per-model periods mean more than one reset time to track.
Kimi Code
Kimi Code usage limit explained
A coding-plan quota with short windows, not a per-message count.
Warp
When do Warp credits reset?
One pool resets on a date; the other only runs out. Read them separately.
- Codex
Codex reset tracker: the next 5-hour and weekly window
Two countdowns, two behaviours. How to know exactly when Codex lets you work again.
- Codex
How to see remaining Codex tokens
Remaining capacity is reported as a percentage. Here is how to read it and estimate the tokens behind it.
- OpenCode
OpenCode model usage: tokens and cost per model
Provider plus model is the split that explains an OpenCode bill.
- OpenCode
How to use OpenCode: install, models, and usage
From install to first tracked session, with the paths you will need later.
GitHub Copilot
Copilot token usage: what actually counts
Why Copilot reports credits rather than tokens, and how to estimate the token side.
Kilo Code
Kilo Code credit balance
The balance lives in the Kilo account, and the account context decides which number is real.
- Comparisons
Best value AI coding subscription
Value is finished work per dollar, measured over a full billing cycle.
- Start here
Codex pricing for every ChatGPT plan
Codex access follows your ChatGPT plan, with windows and credits that vary.
- Start here
Claude Code pricing and plans
One product, four buying routes, two windows that apply to all of them.
- Comparisons
Best free AI usage tracker
Free means open source and local, not a trial with a countdown.
- Claude Code
Claude token usage by model
One total hides the useful fact: which model spent it.
- Codex
Codex CLI /usage command: what it shows and what it hides
The slash commands that answer usage questions, and the four gaps they leave.
- OpenCode
OpenCode limit reset: when your usage windows refill
Three windows, three clocks. The reset you wait for depends on the limit you hit.
- OpenCode
OpenCode free models: limits, resets, and trade-offs
Free models are free until they are not. The limit lives with the provider, not the CLI.
- OpenCode
OpenCode usage dashboard
Four panels that make OpenCode usage readable, and the limits of local data.
- Start here
What is an AI usage tracker?
A reader for data your tools already write, plus the arithmetic.
- Comparisons
Claude $100 plan vs Codex $100 plan
Same price band, different metering. Compare blocked days, not price sheets.
- Comparisons
Copilot vs Cursor vs Claude Code pricing
Premium requests, requests, and token windows: three different meters.
- Start here
AI credit consumption explained
Credits are a billing unit with a model-dependent exchange rate.
- Comparisons
Token usage tracker comparison
Coverage, source, history, limits: the four axes that decide the choice.
- Start here
AI coding usage for teams
Collect aggregates, report by team, and give developers their own numbers first.
- Start here
How is AI coding usage measured?
Four units, four meanings. Converting between them is an estimate.
- Comparisons
Free AI coding plans compared
Free tiers are for evaluation. Their real cost is the work you cannot do.
- Start here
Context window vs usage window
One limits a single request, the other limits a period of work.
- Start here
AI budget guardrails for coding agents
Four guardrails, in order of how much they save per unit of friction.
- Privacy
Local vs provider-reported usage data
Local data is detailed; provider data is authoritative. You need both.
- Start here
Why AI usage numbers do not match
Different scopes, fields, and rounding. Neither number is wrong.
- Start here
AI coding subscription ROI
ROI is finished work per dollar, including the cost of being blocked.
- Start here
AI coding cost per developer
Divide spend by active developers, then segment by workload.
- Start here
AI usage glossary: tokens, credits, quotas, windows
The ten terms that explain almost every usage question.
- Codex
Reduce Codex token usage: 9 tactics that work
Where Codex tokens actually go, and the changes that reduce them without reducing output.
- Codex
Codex usage history: how far back it goes
History is local, per session, and only as durable as the files you keep.
- Start here
AI coding cost analysis for enterprises
Three cost lines, one report, and a measurement method that survives audit.
- Comparisons
UsageAtlas vs TokenTracker
Token counts are the easy half. Limits and resets are the useful half.
- Codex
Codex usage not showing: 6 causes and fixes
A short diagnostic order for when Codex writes no usage data, or the app cannot see it.
- Codex
Codex cost per project: how to measure it
Group sessions by working directory, apply the rates you actually pay, and compare projects fairly.
- Codex
Codex vs Claude Code: cost per task
A fair cost comparison needs the same task, the same measurement, and the same rate basis.
- Codex
Codex usage alerts: warn before you hit the limit
One notification per window per reset cycle, at the threshold you choose.
- Codex
How Codex counts tokens: input, cached, and output
The four fields, one worked example, and the arithmetic that avoids double counting.
- OpenCode
OpenCode stats command: what it shows and what it misses
The CLI snapshot, its blind spots, and the one query that fills them.
- OpenCode
OpenCode vs Claude Code: usage and cost
Two different models of usage: open multi-provider client versus a first-party plan.
- OpenCode
OpenCode usage alerts
Local spend thresholds work today; provider windows need a connected account.
- OpenCode
OpenCode usage not showing: 5 fixes
A short diagnostic order for missing OpenCode usage data.
GitHub Copilot
AI credits vs premium requests: what changed
A count became a currency. Why the old mental model breaks.
GitHub Copilot
Copilot usage report for organizations
What the org report contains, what it omits, and how to present it honestly.
GitHub Copilot
Copilot limits by plan: Free, Pro, Business, Enterprise
What changes between plans, and the only place the current number is published.
- Claude Code
Claude Pro vs Max vs Ultra for Claude Code
Three tiers, one window structure, and a break-even test for each.
- Claude Code
Claude reset tracker
Two countdowns with different behaviour, in one view.
- Claude Code
Cache write tokens explained
The upfront cost that makes later turns cheaper, and when it is wasted.
- Cursor
Cursor plans compared: Pro, Ultra, Teams
What changes between tiers, and the test that tells you if Ultra pays.
- Cursor
Cursor usage per model: Auto vs premium
Auto routes for you; premium models bill by model. That is the whole split.
Antigravity CLIBaseline model quota reached: what to do
One model is out, not the account. Check the other quotas before stopping.
- Comparisons
Google AI Plus and Antigravity limits
A consumer subscription and a coding agent quota are different layers.
- Start here
Usage-based pricing for AI coding tools
You pay for what you use; the risk is that agents use a lot.
- OpenCode
OpenCode token usage: where the numbers are stored
Every assistant message carries its own token counts. Here is where they live and what each field covers.
- OpenCode
Where OpenCode stores its database (and how to find it)
One command prints the real path; the defaults differ by OS and install method.
- OpenCode
OpenCode session history: find and read past sessions
The TUI lists recent sessions; the database holds every one of them, with messages attached.
- Codex
Codex token usage by model
Model-level token totals, why the strongest model is not always the biggest line, and how to compare.
- Codex
Codex API vs subscription: the break-even math
A repeatable break-even calculation, plus the constraints a price table ignores.
- Codex
Codex usage on Mac
The macOS paths, the menu bar view, and the two settings that break tracking.
- Codex
Codex free tier limits
What free access gives you, what it does not, and how to measure the gap.
- Codex
Cached input tokens in Codex: what they save
How prompt caching works in Codex, why a stable prefix matters, and how to see the saving.
- OpenCode
OpenCode usage on Mac
macOS paths for OpenCode, the permission prompt, and a menu bar view.
- OpenCode
OpenCode vs Codex: which is cheaper per task
A metered multi-provider client against a plan with two windows.
- OpenCode
OpenCode token tracker for the CLI
Three terminal-first ways to see OpenCode token usage, from one command to one script.
- OpenCode
OpenCode cost per session: how to calculate it
The formula, the cache trap, and how to compare sessions of different lengths.
- OpenCode
OpenCode usage across multiple providers
One client, many bills. How to keep provider totals honest.
- OpenCode
OpenCode Go: how to check usage
Local tokens are visible; plan limits live with the plan.
GitHub Copilot
GitHub Copilot usage API
The endpoints, the token scope, and a minimal request you can test.
GitHub Copilot
Copilot premium request overage: costs and caps
Whether you get billed, blocked, or both, and how to set the guardrail first.
GitHub Copilot
Copilot usage history
Built-in history is a window; a durable record needs exports.
GitHub Copilot
Copilot usage stats
Seven metrics, what they mean, and the two that usually mislead.
GitHub Copilot
Copilot monthly limit reached: what to do
A diagnostic order for a blocked account, from the unit label to the billing setting.
GitHub Copilot
Copilot token limit per month
There is no token quota. There is a priced allowance, and that is what to plan around.
- Claude Code
Claude Code cost per project
The directory already encodes the project. Here is how to turn it into a number.
- Claude Code
Claude Code limits per developer
Per-seat allowances, shared versus individual windows, and the reporting gap.
- Claude Code
Claude usage in chat vs Claude Code
One allowance, two workloads, and a simple way to attribute the drain.
- Claude Code
Claude daily vs weekly limit
There is no daily reset. There is a rolling window and a weekly budget.
- Claude Code
Claude Code usage alerts
One alert per window per reset cycle, at a threshold you choose.
- Claude Code
Opus vs Sonnet cost per task
The stronger model can be cheaper if it finishes in fewer turns.
- Cursor
Cursor analytics and telemetry
Product telemetry, team analytics, and usage accounting are three separate things.
- Cursor
Cursor expense tracking for teams
Seats, included usage, and on-demand spend belong in three separate columns.
Antigravity CLIAntigravity starter quota: what you get
A trial allowance for evaluation, measured per model like the paid quotas.
Antigravity CLIAntigravity AI credits vs model quota
Quota limits a model; credits pay for usage the quota does not cover.
Antigravity CLIAntigravity weekly limit
Quota periods are per model; not every one is weekly.
Antigravity CLIAntigravity CLI conversation history
Sessions are local files. Here is how to find them and what they hold.
Kimi Code
z.ai coding plan limits
A short-window quota for the coding plan, separate from API keys.
Qoder
Qoder free trial: credits and limits
Trial credits are for evaluation. Measure one task before deciding.
Qoder
Qoder usage dashboard
Personal or shared, the balance is only meaningful with the account context.
Warp
Out of Warp credits: what happens next
AI features stop; the terminal does not. Then it is a purchase or a wait.
Factory / Droid
Factory Standard vs Core limits
The difference is allowance and model access, not how a request is metered.
Amp
Amp agent credits and workspace balance
Personal and workspace balances are separate pools with separate owners.
- Codex
Codex usage graph: see your token trend
What a useful Codex chart shows, and the three shapes that matter.
- Codex
Codex statusline: show usage in your terminal
Where Codex exposes usage, and how to surface it without slowing every shell prompt.
- Comparisons
UsageAtlas vs SigRank
Two local readers, different emphasis: ranking signals versus plan history.
- Codex
Codex tokens per task: benchmarks by task type
Task-level estimates, the variables that move them, and a simple way to measure your own.
- Codex
Codex usage on Linux
Linux paths for Codex, the package options, and the container trap that hides usage.
- OpenCode
OpenCode usage on Linux
Linux paths for OpenCode and two ways to read them: the app or SQL.
- OpenCode
OpenCode info.cost: reading the message field
One field, three possible states, and what to do in each.
- OpenCode
How many tokens does an OpenCode request use?
Request size is a context question, not a client question.
- OpenCode
Export OpenCode usage to CSV
A clean export schema, the SQL to produce it, and the privacy check before sharing.
- Cursor
Is there a Cursor usage API?
There is a team admin API and a model API, and they answer different questions.
Kimi Code
Kimi API rate limits
Coding-plan windows and API rate limits are separate systems.
Kimi Code
Kimi Code cost per session
Plan price divided by sessions is the number that matters, not price per token.
Kimi Code
z.ai vs Claude Code cost
A cheaper per-token provider against a first-party plan with windows.
Pi / OMP
Oh My Pi sessions: list and read them
Sessions are local files with a predictable layout. Here is how to navigate them.
- Comparisons
Kimi Code vs Claude Code cost
Different metering, same test: cost per task that actually finished.
- Comparisons
Warp vs Cursor cost
Warp meters credits; Cursor meters requests and model usage. Compare per task.
GitHub Copilot
Copilot usage per developer: how to measure each seat
Seat data and consumption data live in different reports. Joining them is the only way to see real per-developer usage.
GitHub Copilot
Copilot vs Cursor: how usage and cost really compare
Two different billing units, two different failure modes. What to measure before you call one cheaper.
- Codex
Codex usage for teams: per-developer reporting
Per-seat reporting has two halves: local consumption and organization billing.
- Codex
Codex login status error: fix it fast
A short diagnostic path for auth errors that look like usage problems.
- Codex
Codex Business plan usage limits
Per-seat windows, admin visibility, and the questions to ask before a team rollout.
GitHub Copilot
Copilot analytics for teams
A team analytics setup that answers budget questions without ranking people.
GitHub Copilot
Copilot usage in JetBrains
The plugin view, its limits, and the account page behind it.
GitHub Copilot
Copilot Business seat usage
Assigned, active, and consuming are three different seat counts.
GitHub Copilot
Copilot data usage setting explained
Three different settings people confuse: telemetry, code retention, and model training.
GitHub Copilot
Copilot chat vs agent usage
One turn versus a loop of turns, and what that does to your allowance.
- Claude Code
Claude Code Team plan limits
Per-seat capacity with organization controls, not a shared pool.
- Cursor
Cursor usage per developer
Admin console for seats and spend, local data for project detail.
- Codex
Codex computer use: token usage explained
Why screen-driven tasks cost more, and the limits that keep them predictable.
- Codex
gpt-5-codex vs mini: tokens, speed, and cost
A decision table rather than a benchmark claim: when to use each model.
- Claude Code
Claude Code 7x tokens: what changed
A multiplier changes capacity, not your habits. Verify it against real sessions.
- Claude Code
Claude desktop app usage tracker
Chat usage and Code usage share a plan but not a data source.
- Claude Code
Claude Code Enterprise usage
More control and reporting, the same two windows.
- Claude Code
Cheaper Claude Code alternatives
Five ways to cut cost without pretending the cheap option is identical.
- Claude Code
Export Claude usage history
A clean export schema from the logs, without prompts or file contents.
- Comparisons
The best AI coding usage trackers in 2026, compared
Terminal reports, menu bar meters, and desktop dashboards, sorted by what you need.
- Comparisons
CodexBar for Windows: what to use instead
No official Windows build exists. These are the options that actually run there.
- Claude Code
Claude usage limit reached? What to do next
Read the message, find the reset time, then choose to wait, pay, or switch.
- Claude Code
How to check Claude Code usage
Three places to look, and what each one leaves out.
- Claude Code
How much does Claude Code cost in 2026?
Plan prices, API billing, and a way to price your own usage before you pick.
- Codex
Codex usage limit reached? What to do next
Find the window, find the reset, then pick credits, waiting, or a different tool.
- Comparisons
ccusage alternative with a desktop dashboard
Keep ccusage for the terminal. Add a window when you want limits, alerts, and charts.
- Comparisons
CodexBar vs UsageAtlas: menu bar meter or usage dashboard
Two free, open-source trackers with different jobs. Pick by platform and by how much history you want.
- Windows
Claude usage tracker for Windows
A Claude usage widget for the Windows tray, with history behind it.
- Claude Code
See Claude Code token usage by day, project, and model
Where your Claude Code tokens went, broken down the way you want to read them.
- Claude Code
When does the Claude weekly limit reset?
Not midnight, not Monday. A time tied to your account, and how to find it.
- Claude Code
Claude 5-hour session limit vs weekly limit, explained
Two meters run at once. The short one stops your afternoon, the long one your week.
- Claude Code
Claude Pro vs Max: usage limits for Claude Code
The multipliers are per session. Measure your weekly use before paying for more.
- Claude Code
How to reduce Claude Code token usage
Changes that cut Claude Code token use, and how to see the effect.
- Codex
When does Codex usage reset?
Two windows, two reset times, and where to see each one.
- Cursor
How to check Cursor usage in 2026
New plans, legacy plans, and two dashboards. Here is which one is yours.
- Start here
Claude Code vs Codex: usage limits and cost compared
Same price tiers, different meters. Run both and let your history decide.
- Start here
Get an alert before you hit an AI coding limit
One notification at the percentage you choose, before the limit stops your agent.
- Start here
AI usage tracker for Linux
The same usage dashboard on Linux, installed from an AppImage, .deb, or .rpm.
- Start here
Input, output, and cache tokens explained
Why your token total looks huge, and which part of it actually costs money.
- Claude Code
Claude Code usage credits, formerly extra usage, explained
Pay as you go past the limit, with a cap you set. Extra usage got a new name.
- Claude Code
Claude Code deletes history after 30 days. Here is how to keep it
The default cleanup erases last month. Two ways to keep the record.
- Claude Code
Claude Code with an API key or a subscription: which costs less?
Flat fee with limits, or pay per token with none. Your token history decides.
- Claude Code
Is Claude Max worth it? Check your usage first
The answer is in your limit history, not in someone else's review.
- Codex
Codex usage limits on ChatGPT Plus, Pro, and Business
What each ChatGPT plan gets in Codex, and why the numbers keep moving.
- Codex
Codex credits explained
Credits carry you past the limit. Know your balance before a long run.
- Windows
Codex usage tracker for Windows
Codex limits in the Windows tray, with local token history behind them.
- Cursor
Cursor usage limit reached? What to do next
Which pool ran dry, when it refills, and what extra usage will cost.
- Start here
AI usage tracker for Mac
Menu bar presence, reset times, and token history for every tool on your Mac.
- Start here
Add AI coding stats to your GitHub README
Copy one Markdown line and your README shows your Claude Code, Codex, and Cursor activity.
- Start here
AI Coding Wrapped: your year in Claude Code, Codex, and Cursor
A year-in-review page for your AI coding, built from the usage you saved.
- Start here
AI coding leaderboard for tokens and streaks
Public profiles ranked by tokens and streaks, with filters by tool, model, and period.
- Start here
How many tokens does AI coding use per day?
Averages from strangers will not help you. Your own baseline will.
- Start here
Track AI coding usage and cost per project
Find the repository that ate your week, and estimate what it cost.
- Comparisons
OpenUsage for Windows and Linux: a free alternative
OpenUsage stays on the Mac. This is what to run on your Windows or Linux machine.
- Comparisons
ClaudeBar alternative for Windows and Linux
ClaudeBar needs macOS 15. Here is a free tracker for the machines it does not cover.
- Comparisons
Claude Code Usage Monitor alternative with a desktop app
From a terminal burn-rate meter to a desktop view of every tool you use.
- Comparisons
ccstatusline or a usage dashboard for Claude Code?
A status line for the session in front of you, a dashboard for everything else.
- Comparisons
cursor-stats is archived. Track Cursor usage this way instead
The old status bar extension stopped. Your Cursor usage still needs a home.
- Comparisons
Free SessionWatcher alternative for Windows, macOS, and Linux
The same job, no license fee, and a Linux build.
- Comparisons
CodeBurn vs UsageAtlas: cost optimizer or usage dashboard
Pick CodeBurn to cut spend, UsageAtlas to watch limits and keep a long record.
- Claude Code
Claude Code /cost and /stats now open /usage
The old commands still work. They open a different screen now.
- Claude Code
Claude "Reset for free": how limit resets work
A button that sometimes shows up on the web. It never shows up in the terminal.
- Codex
Is ChatGPT Pro worth it for Codex?
Count how often Plus stops you. That number answers the question.
- Cursor
Cursor vs Claude Code: cost and usage limits
Different meters, similar price tiers. Track both for a week before you choose.
Antigravity CLIGemini CLI usage limits after the switch to Antigravity CLI
The daily request quota moved for most people. Here is what replaced it.
Antigravity CLIAntigravity quota: how limits refresh and where to check them
Five-hour refreshes, a weekly ceiling, and the quota panel in the CLI.
- Start here
Which AI coding subscription is cheapest for you?
List prices tell you little. A week of your own usage tells you a lot.
- Start here
Compare your AI coding usage with a friend
Two public profiles side by side, from tokens to favorite models.
- Comparisons
Track Claude Code usage without a proxy
You do not need a server between Claude Code and Anthropic to see your usage.
- Comparisons
WakaTime AI dashboard vs UsageAtlas
One measures code written with AI. The other measures the plan you are burning through.
- Comparisons
Do you need Langfuse or Helicone to track Claude Code usage?
Observability platforms are for products you ship. Your own coding agent needs less.
GitHub Copilot
How to track GitHub Copilot usage and AI credits
Check Copilot usage in GitHub, then track reported credits and quotas in UsageAtlas, a free, open-source desktop app.
Warp
How to track Warp AI usage and credits
Connect Warp to UsageAtlas to see monthly credits, add-on balances, and reported resets. Free and open source on your desktop.
Kimi Code
How to track Kimi Code usage
See Kimi Code usage windows in UsageAtlas. Connect a Coding API key, choose your region, and check reported quota and reset times.
Kilo Code
How to track Kilo Code usage and credit balance
Track Kilo Code credits and Kilo Pass in free, open-source UsageAtlas. Choose your personal or organization account for the right balance.
Factory / Droid
How to track Factory Droid usage
Connect Factory to UsageAtlas for reported Standard and Core limits, balances, and resets. A free, open-source tracker for Droid users.
Amp
How to track Amp usage and credits
Track reported Amp allowances, credit balances, and Orb hours in UsageAtlas. Connect an access token to the free, open-source desktop app.
Qoder
How to track Qoder usage and credits
Check Qoder credits, then connect your account to UsageAtlas for personal and shared balances. Free, open-source tracking on your desktop.
Antigravity CLIHow to track Antigravity CLI token usage
Track supported Antigravity CLI tokens, models, and local session history with free, open-source UsageAtlas. Learn which files it can read.
Pi / OMP
How to track Pi and Oh My Pi token usage
See Pi and Oh My Pi tokens, models, sessions, and available cost estimates in UsageAtlas, a free, open-source desktop usage tracker.
Muse Code
How to track Muse Code token usage
Track Muse Code tokens, models, and session history from supported local logs with UsageAtlas. Free and open source, with no account needed.
GitHub Copilot
GitHub Copilot AI credits and premium request usage explained
Find Copilot AI credit usage and understand older premium-request counters. UsageAtlas displays the quotas your account actually reports.
Kimi Code
Kimi Code usage limit reached? Check your quota reset
Check which Kimi Code window has run out, find its reported reset, and monitor supported limits in the free UsageAtlas desktop app.
Kilo Code
Kilo Code credits and Kilo Pass: check the right balance
Understand personal credits, organization balances, and Kilo Pass allowances. See the selected account's reported balance in UsageAtlas.
Warp
Warp AI credits: how to check usage and remaining balance
Find Warp's credit usage and keep monthly and add-on pools visible in UsageAtlas. Credit counts are different from model requests or tokens.
Factory / Droid
Factory Droid usage limits and reset times explained
Read Factory's reported Standard and Core windows without mixing their allowances. Keep available limits and resets in view with UsageAtlas.
Amp
Amp credits and Orb hours: how to check remaining usage
Agent dollars, Orb hours, and workspace credits measure different things. UsageAtlas keeps Amp's reported pools separate on your desktop.
Qoder
Qoder credits: check personal and shared usage
Find Qoder's used and remaining credits, understand shared pools, and monitor the available counters in free, open-source UsageAtlas.
- Start here
AI usage not showing in UsageAtlas? Check these settings
Fix missing providers, expired connections, and unreadable local history. Learn why a credit-only integration does not show token history.
- Claude Code
Claude Code usage history on your desktop
/usage shows now. History is what tells you how you got here.
- Codex
How to check Codex usage
/status for a glance. A desktop dashboard if you also want history and the rest of your tools.
- OpenCode
How to see OpenCode usage
OpenCode already stored the numbers. This is how you look at them.
- Codex
Codex usage dashboard
Codex usage stats on a desktop dashboard, not a billing page and not a status line.
- OpenCode
OpenCode usage on Windows
OpenCode usage for Windows, read from the local database already on the machine.
- Claude Code
See Claude usage last month
Last month of Claude Code usage, without reconstructing it from a live quota.
- Codex
View Codex token usage
Token counts for Codex work you already ran, plotted against your own days.
- OpenCode
OpenCode cost tracking
OpenCode often sits in front of several model bills. This is how you still see a cost picture.
- Windows
AI usage tracker for Windows
The same local dashboard on Windows 10 and Windows 11.
- Claude Code
Claude Code limits and reset times
The current Claude Code window, how much is left, and when it resets.
- Codex
How to check the Codex context window
Context window and usage window are different numbers. Here is how to read both.
- Cursor
Cursor plan usage
Plan usage lives in Cursor. This puts it on a dashboard you can check before a long agent run.
- Start here
Monitoring AI coding usage across tools
Watch usage across tools, with available cost estimates and supported quota alerts.
- Claude Code
Track Claude Code and Codex usage together
One running total when the day is split across Claude Code and Codex.
- Start here
AI rate limits and reset times in one view
Remaining capacity and reset clocks, without guessing from a status line.
- Codex
ccusage for Codex, and a desktop alternative
Keep ccusage if the terminal is enough. Use UsageAtlas when you want a dashboard and more than one tool.
- Privacy
Anonymous usage stats, without sending your prompts
Local tracking, optional free cloud save, and a separate anonymous install count.
- Other tools
How to check z.ai usage
z.ai has its own usage page. UsageAtlas covers a different set of coding tools.
- Start here
Free AI usage tracker for 14 coding providers
One local dashboard for tokens, cost, limits, and history across the tools you already use.
- Claude Code
Track Claude Code usage, limits, cost, and history
Keep Claude Code capacity, cost, and history in view while you work.
- Codex
Codex usage tracker for windows, tokens, and credits
Codex session windows, weekly limits, credits, and local token history in one desktop view.
- OpenCode
OpenCode usage tracker for tokens, models, and sessions
See OpenCode token usage, cost, and sessions in the same dashboard as your other tools.
- Start here
AI coding cost tracker for tokens and estimated spend
Turn tokens and models into an estimated cost picture. Confirm the invoice with the provider.
- Cursor
Cursor usage tracker for plan limits, Auto, and API
Keep Cursor plan, Auto + Composer, and API usage next to your other AI coding tools.
