Grok and Perplexity limits, tracked live
Grok caps each request kind on its own rolling window, and Perplexity splits a daily allowance from Deep Research. tokenkarma shows how much of each you have used and when it resets, so neither one stops you cold.
Grok (from xAI) does not run one quota. It limits each request kind, Auto, Fast, Expert, Heavy, DeepSearch, plus image and video, on its own rolling window. Image generation has used a roughly 2-hour window. SuperGrok and Premium tiers raise the caps.
Perplexity Pro gives a large daily search allowance plus a separate cap on Deep Research, around 500 per day. In mid-May 2026 Perplexity quietly reduced limits, and neither provider shows a clear live counter. tokenkarma tracks both windows and warns you first.
How Grok usage limits work
Grok is the part most usage trackers ignore, because xAI does not expose one simple number. Instead, each request kind draws from its own pool on its own rolling window, so you can be blocked on one mode while another still works.
Quotas are per request kind
The same Grok session can meter Auto, Fast, Expert, Heavy and DeepSearch separately, and media features (image and video) are metered separately again. Hitting your image cap does not touch your text quota, and the reverse is also true. Each kind resets on its own schedule.
The image window
Grok image generation has run on a rolling window of roughly 2 hours, which is why a "grok image limit reset" search is so common. Media features have been throttled at times, so the effective reset can arrive slower than the nominal window suggests.
SuperGrok and Premium tiers
Higher tiers (SuperGrok, X Premium and Premium Plus) raise the caps on the heavier request kinds and on media. The structure stays the same: more headroom per kind, not one shared bucket. For the engineering view, see the hidden rate limits of every major AI API.
How Perplexity usage limits work
Perplexity is the other provider trackers skip, partly because the brand term is polluted by SEO rank-tracker tools. Its limits are simpler than Grok's but split across two pools, and the recent tightening caught a lot of Pro users off guard.
Daily search vs Deep Research
Perplexity Pro gives a large daily search allowance and a separate cap on Deep Research, which has sat around 500 per day. The pools are independent: running out of Deep Research does not stop ordinary Pro searches, and vice versa.
The mid-May 2026 reduction
In mid-May 2026, Perplexity quietly reduced limits. Pro users reported hitting a weekly quota at only 10 to 20 queries a day, with little warning and no clear live counter. This mirrors a broader 2026 trend of providers tightening allowances, covered in the June 2026 subscription nerfs.
When Perplexity resets
Perplexity allowances reset on a daily cycle, with Deep Research tracked as its own pool. Because there is no obvious counter in the interface, the wall tends to feel sudden. tokenkarma shows the remaining queries and the reset time for both pools.
When do Grok and Perplexity reset?
The two providers reset on very different mechanics, which is why a single mental model never works. Grok windows are short and per-kind. Perplexity is daily with a separate Deep Research pool. Here is the shape of both, side by side.
| Provider | Limit | Window | Resets |
|---|---|---|---|
| Grok | Text request kinds (Auto, Fast, Expert, Heavy, DeepSearch) | Rolling, per kind | Per kind, independently |
| Grok | Image generation | Rolling, roughly 2 hours | About 2h after the requests that filled it |
| Grok | SuperGrok / Premium | Same structure, higher caps | Per kind, with more headroom |
| Perplexity | Pro daily search | Daily allowance | On a daily cycle |
| Perplexity | Deep Research | Separate cap, around 500/day | Daily, independent of search |
Caps and windows above are observed behaviour as of June 2026 and can change at any time. tokenkarma is not affiliated with xAI or Perplexity.
How to track Grok and Perplexity usage
Neither provider gives you a clear day-to-day view. Grok hides its per-kind pools, and Perplexity shows almost nothing until you are already blocked. tokenkarma closes both gaps. For each tracked limit it shows:
- how much is used and how much is left,
- the reset time for each window or pool,
- an alert before you hit the wall,
- Grok and Perplexity next to every other AI you pay for.
These are long-tail providers that no dedicated tracker covers well, which is exactly the wedge: see them in the same place as Claude, ChatGPT, Gemini and Cursor. For the cost angle, read how to track AI spending across multiple providers.
Track them next to every other AI you pay for
Most people who pay for Grok or Perplexity also pay for Claude, ChatGPT, Gemini or Cursor, each with its own limits and its own blind spots. tokenkarma puts them all in one view, so you can move work to whichever model still has room. See the full AI usage tracker or the side-by-side limit comparison.
Frequently asked questions
What are Grok’s usage limits?
Grok (from xAI) does not run one single quota. It limits each request kind separately, for example Auto, Fast, Expert, Heavy, DeepSearch, plus image and video, each on its own rolling window. Image generation has used a roughly 2-hour window. SuperGrok and Premium tiers raise these caps. Details are not fully published and can change.
When does my Grok image limit reset?
Grok image generation has run on a rolling window of roughly 2 hours, separate from the text quotas. The window reopens about 2 hours after the requests that filled it, not on a fixed daily clock. Media features have been throttled at times, so the effective reset can feel slower than expected.
What is the Perplexity daily limit?
Perplexity Pro gives a large daily search allowance plus a separate cap on Deep Research, which has sat around 500 per day. The two pools are independent: you can still run searches after exhausting Deep Research. In mid-May 2026 Perplexity quietly reduced limits, frustrating Pro users.
When does Perplexity reset?
Perplexity allowances reset on a daily cycle, with Deep Research tracked as its own pool. After the mid-May 2026 reduction, some Pro users reported hitting a weekly ceiling at only 10 to 20 queries a day. Perplexity does not surface a clear live counter, which is why the wall feels sudden.
Is there a usage tracker for Grok and Perplexity?
Yes. tokenkarma tracks Grok and Perplexity in the same place as Claude, ChatGPT, Gemini and Cursor. These are long-tail providers that no dedicated tracker covers well, so most heavy users have no way to see how much is left until they hit the wall.
Is tokenkarma affiliated with xAI or Perplexity?
No. tokenkarma is independent and not affiliated with xAI or Perplexity. The limits described here reflect publicly observed behaviour as of June 2026 and can change at any time.
Stop guessing your Grok and Perplexity limits
tokenkarma tracks your Grok request-kind quotas and your Perplexity daily and Deep Research pools live, with reset times and alerts before you hit a wall.
Free to start. Works in your browser and on your Mac.
Not affiliated with xAI or Perplexity. Limit mechanics reflect publicly observed behaviour as of June 2026 and can change at any time.