AI with no limits
Short answer: paid models have no monthly token limit. The guest tier caps at 30 messages an hour, and free models on a paid account cap at 10M tokens a month (1M on pay-as-you-go). Plans start at $20 a month, or $0 pay-as-you-go.
Almost every “unlimited AI” page is describing a limit it hasn’t mentioned yet. So here are the numbers, including where they bite.
The limits, exactly
| Limit | |
|---|---|
| Guest, no account | 30 messages an hour on Llama 3.3 70b Instruct, replies up to 2,048 tokens |
| Pay as you go, free models | 1M tokens a month |
| Pay as you go, paid models | No monthly token limit; pay per use from the balance |
| Casual, free models | 10M tokens a month |
| Casual, paid models | No monthly token limit |
| Pro, everything | No monthly token limit |
| Concurrency | A queue, with priority by plan: Pro first, Casual next, guests last |
The distinction that matters is free models vs paid models. Free models are subsidised, so they get a token ceiling. Paid models are billed per token from your balance, so a ceiling would be arbitrary: you are already paying for exactly what you use.
There is no per-message cap on paid models and no daily message count to reset.
What “no limit” does still have
Honest list:
- A queue. Pro is served first, Casual next, guests last. Under load a guest waits and a Pro account doesn’t.
- A balance. On pay-as-you-go, the limit is your balance. That is a real limit even if it isn’t a token cap.
- A 2,048-token reply ceiling for guests. Pro accounts are not capped this way.
- No image or video. If a page promises unlimited images, it is not talking about this service.
Why the plans are shaped like this
$20 a month buys $15 of usage at list prices plus uncapped paid models. Pro at $45 adds $35 of usage, the uncensored models and the front of the queue. Pay-as-you-go is $0 a month and no commitment, which is the right plan for occasional use and the wrong one if you chat daily.
Usage beyond the included credit is billed from the balance at the same per-model prices, so nothing stops working: it just starts costing. Per-model prices are on the model list.
Starting without a plan
Guest chat is free with no account and no cookies. When 30 an hour stops being enough, an account is a random 16-digit number, and Pay As You Go needs no monthly fee. New accounts get $1 of starter credit while this month’s promo budget lasts.
Common questions
Is there really no message limit?
On paid models, there is no monthly token limit and no daily message count. Guests are capped at 30 messages an hour.
What is the guest limit?
30 messages an hour, on Llama 3.3 70b Instruct, with replies up to 2,048 tokens. It resets hourly.
Do free models have a token limit on a paid plan?
Yes: 10M a month on Casual and Pro, 1M on pay-as-you-go. Paid models have none.
What happens when I run out of included usage?
Usage past the included credit is billed from your balance at list price. The plan does not stop.
Does a limit reset daily?
Only the guest hourly cap resets, hourly. Free-model token allowances are monthly.
Can I get an account with no email to hold the limit?
Yes. The account is a number, with no email, phone or password.
Last updated 2026-11-02. Plan prices and model lists on this page are read live from the site.
Try it now: guest chat needs no account, no email and no cookies. When you want more models, an account is just a random number.
See the plans See pricing