Free
Plan 01
$0forever
No account. No card.
Lipi sits above the chat box and shows your 5-hour session and weekly usage, with reset countdowns and alerts. It also brings a prompt library, chat export and hand-off to other AI tools.
Saves the full conversation with artifacts as Markdown.
Right now
You are at 31% this session. Resets at 3:50 AM.
At this pace: ~91% of your weekly limit by the reset.
Close to your limit
Plenty left. This is a good time to start a big task.
Insights grow as you use Claude. Cards fill in on their own.
Nothing starred yet. Star a prompt in History to keep it here.
Guides
Ask for bullets, a table or a short paragraph. A clear format saves a follow-up message.
Number the parts. Each step gets a clearer answer than one long request.
Writing tips and model recommendations. Always free.
Demo with sample numbers. Alerts at 75%, 90% and 100% are yours to switch on. The real Lipi reads your own usage.
Blue means plenty left. Orange means you are getting close. Red means you are nearly out. The toolbar icon shows the same color, so you can check without opening anything.
Pick a feature and try its mock-up. Everything here uses sample data.
The toolbar icon shows your session percent, in the same color.
The strip above the message box
It shows the same numbers while you chat, and “~N msgs left” is an estimate worked out from your own usage.
Reads the same usage numbers Claude shows you, using your own signed-in session.
FreeArtifacts are zipped when the chat has any.
# Q3 report draft > Exported with Lipi on Oct 2, 2026 ## You Draft the Q3 report summary for the leadership team ## Claude Here is a first draft with three headline results and two risks. …
Exports are built in your browser. Artifacts are bundled only when you ask for them.
Free614 templates in 16 categories. 80 are free. A few examples:
Bug triage
Get a step-by-step root cause analysis, ranked hypotheses, and a fix, not just a guess.
Refactor plan
A phased, low-risk refactoring roadmap with tests and rollback steps for each stage.
Write documentationPro
README, API docs, or inline comments, written for the developer who has to read it later.
Pick a template to see what happens.
A template goes into the message box so you can edit it before sending.
Free: 80 templates. Pro: all 614.Check a draft against common prompt-writing rules before you send it. Pick a sample draft:
Score
10 / 10
These are real checks from the extension. The checklist reads English prompts only for now. Free: 10 runs. Pro: unlimited.
A quick check, not a verdict. You stay in charge of the final wording.
Free: 10 runs. Pro: unlimited.Take the conversation to another AI tool. Lipi fills the message box or copies the chat. It never presses send.
ChatGPT can be filled in for you. Chrome asks for a one-time permission the first time; after that hand-offs are instant.
## Continuing a conversation from Claude I was working with Claude (Anthropic) and I’m moving this conversation to you. Here’s the context: **Topic:** Q3 report draft **Length so far:** 6 messages, 1 artifact **How it started:** Draft the Q3 report summary for the leadership team --- Last 5 messages --- **Claude:** Here is a first draft with three headline results and two risks. **Me:** Make the risks section shorter and add next steps **Claude:** Updated. The risks are now two lines each, with a next-steps list. **Me:** Turn it into a one-page report **Claude:** Done. The one-page report is in the artifact. --- Artifacts --- **One-page report** (html) Please pick up from here.
Lipi never sends the message for you. You review it in the other tool first.
ProStart a fresh Claude chat that carries the conversation up to the message you choose. Useful when a long chat gets heavy.
Estimated context used: ~2%
Continuing an earlier conversation — here’s what we discussed so far. Please pick up from here. --- You: Draft the Q3 report summary for the leadership team Claude: Here is a first draft with three headline results and two risks. You: Make the risks section shorter and add next steps Claude: Updated. The risks are now two lines each, with a next-steps list. You: Turn it into a one-page report Claude: Done. The one-page report is in the artifact. ---
Your original chat is not changed. The fork opens in a new Claude tab, ready for you to send.
ProBuilt from your own usage, on your computer. Pick a sample week:
At this pace: ~59% of your weekly limit by the reset.
Room to spare
Plenty left. This is a good time to start a big task.
Busiest on Tuesday. Quietest on Sunday.
Insights fill in as you use Claude. The weekly forecast needs history from earlier weeks.
FreeSave the accounts you use and move between them in one click. You confirm each switch.
PersonalActive
Session 22% · Weekly 41%
WorkActive
Session 63% · Weekly 77%
Free: 2 saved accounts. Pro: up to 5. Saved sessions stay in your browser on your computer.
You confirm every switch. If a switch fails, Lipi puts your previous session back.
Free: 2 accounts. Pro: up to 5.Desktop notifications before you run out. You choose the thresholds, for the session and the week.
Also in Settings: your own color thresholds and how often Lipi checks. The popup is available in English, Spanish, French, German, Portuguese, Italian, Japanese and Korean.
Alerts appear on your desktop. By default Lipi notifies you at 100% only.
FreePro is for the people who want the prompt tools and hand-off without limits.
Plan 01
$0forever
No account. No card.
$3.99/ month
Billed monthly. Cancel anytime.
Lipi works with your own claude.ai session in your own browser. Here is exactly what that means.
No. Lipi is an independent extension built by Engramium. It is not affiliated with or endorsed by Anthropic.
It is an estimate that Lipi works out from how your own usage moves. Treat it as a guide, not a promise.
Lipi reads the usage data Claude provides. If that changes, a meter may show “No usage data” until Lipi is updated.
Use the contact page at engramium.com. Reviews on the Chrome Web Store help too.