[Verified] Monica's free tier is 40 prompts a day shared by three small models, and Sonnet access starts at US$99 a year

Monica is a browser extension that puts a chat box on top of any web page and lets you pick which AI model answers. It sells one subscription that covers models from Anthropic, Google and OpenAI. That bundle is the whole pitch. We checked the pricing page ourselves, first on 2026-07-31 through a proxy read and again in a browser on 2026-08-09, and re-verified the numbers on 2026-09-12. Our angle is simple: the free tier is a demo, not a plan, and the real question is whether US$99 a year is a fair price for one login that reaches three vendors’ models.

The free tier: one pool of 40, not 40 per model

Sign up, install the extension, and you get 40 prompts a day. A prompt here means one message you send and one reply you get back. The 40 are shared across three models: Claude 4.5 Haiku, Gemini 3.1 Flash-Lite and GPT-5.4 mini. Use 30 on Haiku and you have 10 left for everything else. The counter resets daily.

The vendor page lists the three model names side by side, and it is easy to read that as 40 each. It is not. This is the first gap between the claim and the reality.

The second gap concerns the models people actually want. Claude 4.6 Sonnet, GPT-5.5 and Gemini 3.1 Pro are the “advanced” tier. Free users get what the page calls limited trials of them. No number is given. We could not find one anywhere on the pricing page, so we cannot tell you how many Sonnet replies a free account gets before the wall goes up. Assume very few.

The rest of the free tier is also trial-shaped. Image and video generation are a limited trial. ChatPDF, the feature that lets you ask questions about a PDF, is a limited trial. The AI Agent feature is a time-limited taste. You can sign in on up to 3 devices. Commercial use is allowed and there is no watermark on output, which is better than some rivals.

Everything on the free tier works inside the extension on any page. That part is not gated.

What US$99 a year buys on Pro

Pro is listed at US$99 per year, which the page presents as about US$8.3 a month. That monthly figure is the annual price spread out. We tried the monthly billing toggle on 2026-08-09 and it did nothing, so there is no verified month-to-month price. If you want to pay monthly, we cannot tell you what it costs, and neither can the page.

Pro gives you 5,000 basic-model prompts a month and 200 advanced-model prompts a month. Basic means the same three small models as the free tier. Advanced means Sonnet, GPT-5.5 and Gemini Pro. You also get 1,500 premium credits, a separate points balance Monica meters for heavier features apart from the chat counts, and 5 devices.

The 200 is the number to fix in your head. It is about 6 or 7 advanced replies a day if you spread them evenly, though the quota is monthly so you can burn them in a week. For light use that is fine. For someone who lives in Sonnet all day, it is not.

Max and Ultra: the discount clock and the beta label

Max is US$199 a year, shown as about US$16.6 a month. The page marks this as a limited-time 33% discount from an original price of US$299 a year. No end date is shown. Max gives unlimited use of both basic and advanced models, 4,500 premium credits and 5 devices. Two things to know. It used to be called Unlimited, and the page still labels it as being in a testing phase. Unlimited advanced access is the entire reason to pay for Max, so a beta tag on that promise deserves a raised eyebrow.

Ultra is US$995 a year, shown as about US$82.9 a month, with the page claiming a 17% saving for paying annually. It matches Max on unlimited models but raises premium credits to 20,000 and devices to 10. On the page’s own numbers, that is 5 times the price of Max for 4.4 times the credits. Unless credits are your bottleneck, Ultra is for teams, not individuals.

None of the three paid plans shows a working monthly price. That is the one thing readers keep asking us, and the honest answer is that the page will not tell you.

☀️ AMO, the budget-minded cat

AMO’s question is always the same: is free enough, and if not, what is the cheapest tier that does the job?

Free is enough for exactly one kind of person: someone who wants a quick summary or rewrite a few times a day and does not care which model does it. Claude 4.5 Haiku, Gemini 3.1 Flash-Lite and GPT-5.4 mini are all small, fast models built for that. 40 shared prompts a day covers casual use. The moment you want Sonnet or GPT-5.5 to do real thinking, you hit the unstated trial cap, and there is no way to buy a small top-up. It is free or US$99 a year, with nothing between.

Is the US$99 worth it? The page itself invites the comparison with ChatGPT Plus at US$20 a month, and on that basis Pro at about US$8.3 a month looks like less than half the price for access to three vendors instead of one. That comparison is fair only if 200 advanced prompts a month is enough for you. If you would use ChatGPT Plus heavily, you would blow through 200 Sonnet or GPT-5.5 replies in the first week and spend the rest of the month on the small models. At that point the cheaper number is not a bargain, it is a different product.

AMO’s practical rule: count your advanced-model use for a week before paying. Under 50 a week, Pro at US$99 a year is a genuine saving. Over that, the only sensible Monica plan is Max at US$199 a year, and only while the 33% discount is live. At the original US$299, the gap to buying one vendor’s plan directly narrows enough that the bundle argument weakens. Skip Ultra unless someone else is paying.

🌙 PIMI, the performance-minded cat

PIMI cares whether the thing works, whether the good models are really there, and whether the plan lets you use them.

The extension itself does what it says. It sits on any page, free or paid, and you can send the page text to a model without copying it anywhere. That is the workflow win, and it is real. We did not run timed benchmarks, so we will not make claims about speed.

The model question is where the free tier fails a performance test. The three free models are the smallest each vendor offers. For summarising a page they are adequate. For drafting, coding, or anything needing careful reasoning, you want Claude 4.6 Sonnet, GPT-5.5 or Gemini 3.1 Pro, and free gives you an unspecified handful of trials. A power user cannot evaluate Monica properly on the free tier, because the models worth evaluating are behind the paywall.

Pro fixes the access problem but rations it: 200 advanced prompts a month. The 5,000 basic prompts are effectively unlimited for a human. The design pushes you to draft with the small models and escalate to a big one only when it matters. That is a workable habit for some people and a constant irritation for others.

Max is the plan that actually delivers the performance promise. Unlimited advanced models means you can leave Sonnet or GPT-5.5 selected and forget about counting. The catch is the beta label on the plan. A vendor that calls its unlimited tier a test is reserving the right to change it. Also note that on Pro and Max you get 5 devices, which covers a laptop, a desktop and a phone with room to spare. Ultra’s 10 devices and 20,000 credits are team-scale.

PIMI’s verdict on the models: what you get on Max is the same headline models you would get from each vendor directly, chosen from one menu. That is convenient. It is not faster or smarter than going direct, and the page gives no reason to think it is.

Verdict: who should use it, who should skip it

Use the free tier if you want a page-summary tool that runs on small models and 40 prompts a day is plenty. Do not treat it as a way to try Sonnet or GPT-5.5. The trial cap for those is unstated and you will hit it.

Pay for Pro at US$99 a year if you want one extension that reaches three vendors and your heavy-model use is genuinely occasional. Stay under 200 advanced prompts a month or you will be unhappy.

Pay for Max at US$199 a year if you want unlimited access to the top models and you are comfortable with a plan the vendor still calls a test. Check whether the 33% discount is still showing before you buy, because the page gives no expiry date.

Skip Monica if you need a monthly price. None of the three plans shows one, and the billing toggle did not work when we tried it. Skip Ultra at US$995 a year unless you are buying for a team that burns credits.

The insight the vendor page will not give you: Monica’s value is the bundle, not the free tier. The free tier exists to get the extension installed. The moment you want the models the page advertises in big type, you are choosing between US$99 and US$199 a year, and nothing in between.

Written: 2026-09-12 Price last verified: 2026-09-12 Tool last health-checked: 2026-09-12

🔗 Related on AMPM

• Monica pricing & free tier
• chatgpt vs monica
• claude vs monica

📎 Sources

All prices and quotas in this article come from AMPM’s verified dataset for Monica (last verified 2026-09-12); primary sources:

• https://monica.im/pricing
• https://monica.im

🐾 Meet the cats: AMO and PIMI

AMPM is a Taiwan-based AI-tool pricing watchdog. Our motto: ask before you subscribe. Two cats argue every tool from two sides:

• ☀️ AMO — the budget-minded cat. Always asks: is the free tier enough, and is paying actually worth it?
• 🌙 PIMI — the performance-minded cat. Cares about whether it works well, fast, and gets your job done.

When they are done arguing, you know whether to pay. Every number is checked by us, with the verification date at the end.

🐾 Our sister sites also checked

• Travel insurance and flight-delay coverage checked
• Brokerage fees and account opening checked