To Elon Musk and the SpaceXAI team
Make Grok Great Again
Restore the SuperGrok experience. Fix the limits. Give paying subscribers equal access to the latest Grok models.
Started by @thisisdimm, for the benefit of SuperGrok subscribers
We are SuperGrok subscribers. We pay $30 a month because Grok used to offer something genuinely different: generous chat access and early access to the newest models.
For many of us, that was the reason we subscribed.
Today, that experience is disappearing.
The new usage limits have made SuperGrok significantly more restrictive
In June 2026, SpaceXAI replaced separate daily limits for Chat, Imagine, Voice, and Build with a single shared weekly usage pool.
On paper, a shared pool sounds flexible. In practice, many SuperGrok subscribers experience it as a significant reduction in what their subscription actually provides.
Users are reporting that:
- Expert mode chat can consume around 1–2% of the weekly allowance per prompt, leaving some users with only around 50–100 messages for an entire week
- Normal daily use can push the weekly allowance close to 100% within just two or three days, even without heavy use of Imagine or video
- Chat, Imagine, Voice, and Build all draw from the same allowance, meaning heavy use of one feature can effectively limit access to the others
- Once the weekly allowance is exhausted, fallback access is extremely limited, leaving users with the choice of waiting for the next reset, purchasing additional usage credits, or upgrading to a much more expensive plan
This is a very different experience from the SuperGrok many of us originally subscribed to.
And the comparison with competing services makes the problem harder to ignore.
ChatGPT and Gemini offer usage patterns that can be considerably more generous for everyday conversation. ChatGPT, for example, separates ordinary text chat from limits on more compute-intensive features, while Gemini uses recurring usage resets alongside its broader plan limits.
The exact systems are different, and no serious AI service can promise unlimited access to every advanced model or feature. But users do not compare subscriptions by internal compute accounting. They compare them by a much simpler question:
Can I use the assistant when I need it?
If someone paying $30 a month for SuperGrok repeatedly reaches their limit after only a few days while competing services allow them to continue working, that becomes a real reason to reconsider which AI assistant they use as their primary one.
For many users, switching from Grok to ChatGPT or Gemini does not require changing hardware, learning complicated new software, or making a major financial commitment. It can happen in minutes.
That should matter.
We understand that compute is expensive. We understand that video generation, advanced reasoning, coding agents, and large models cannot offer unlimited usage.
We are not asking for unlimited access.
We are asking for limits that make sense for a $30 subscription and remain competitive with what users can get elsewhere.
Paying chat users should not receive new Grok models last
There is another problem that is becoming just as important: how new Grok models are distributed.
When Grok 4.6 launched on August 12, 2026, it was immediately made available through Cursor, Grok Build, the API, and other developer platforms.
For paying users whose main way of using Grok is the regular chat experience, access can arrive later.
The problem is not one particular waiting period for one particular model. Grok 4.6 could become available to everyone tomorrow, today, or even minutes after this petition is published.
The problem is the principle.
Compared with how quickly competing AI companies increasingly push their newest models into their main consumer products, SuperGrok chat users should not be left wondering when they will receive the model that has already been publicly launched elsewhere.
A developer using Grok through a coding tool should absolutely receive access to the newest models quickly.
But so should a SuperGrok subscriber who pays specifically to use Grok every day through Chat.
We are not asking for Build, Cursor, the API, or developer tools to receive new models later.
We are asking for SuperGrok chat users to receive them at roughly the same time.
If technical reasons make a simultaneous release impossible, the delay should be short, clearly communicated, and predictable — measured in hours or a small number of days, not an open-ended rollout with no clear timeline.
SuperGrok was supposed to give paying users premium access to Grok. Paying subscribers should not feel like they are standing in line behind other products to access the newest generation.
Chat is not a secondary feature.
For many subscribers, Chat is Grok.
And in an increasingly competitive AI market, making your most loyal paying users wait while rival assistants offer generous access to their latest models creates an unnecessary reason for those users to look elsewhere.
What we are asking SpaceXAI to change
- Give Chat its own usage allowance
Chat should not have to compete with Imagine, Voice, video, and Build for the same weekly pool. Give SuperGrok Chat its own generous allowance with a reasonable reset period.
- Protect a minimum allowance for every major feature
Using Build should not leave someone unable to use Chat. Generating images should not make Grok unavailable for the rest of the week. Every major part of the subscription should remain usable.
- Give all SuperGrok subscribers equal access to new Grok models
When a new flagship Grok model launches, paying subscribers should receive access across Chat, Build, Cursor, the API, and other supported products at roughly the same time. There should not be one model generation for developers and another for paying chat users. If a staged rollout is necessary, communicate the expected timeframe clearly.
- Show users what they are actually paying for
A percentage bar is not enough. Show how much usage remains. Explain how different models and features consume the allowance. Tell users clearly which model they are actually using. Make the rules understandable before the limit is reached.
- Do not quietly reduce the value of an existing subscription
When people subscribe based on a certain level of access, major reductions should not be introduced without clear communication. New systems should improve the subscription, not leave users wondering why they are suddenly receiving less for the same price.
- Keep SuperGrok competitive
SuperGrok does not exist in isolation. ChatGPT, Gemini, Claude, and other leading AI assistants are one click away. Their capabilities are improving rapidly, and users can switch their primary assistant far more easily than they can switch most other subscription services. A premium Grok subscription should therefore compete not only on intelligence, personality, or features, but also on something much simpler: how much users can actually use the product they are paying for. The best model in the world has limited value to a subscriber who cannot access it.
We want Grok to succeed
This petition is not against Grok.
It exists because many of us use Grok every day and want to keep using it.
We want SpaceXAI to keep pushing its models forward. We want Build to improve. We want better image generation, better agents, better coding tools, and more capable models.
But that progress should not come at the expense of the people already paying for the core product.
A $30 SuperGrok subscription should feel like a premium subscription.
It should provide enough usage for Grok to genuinely become part of everyday life.
And when SpaceXAI releases its best new model, paying SuperGrok subscribers should not feel that they are among the last people in the Grok ecosystem to receive it.
Because if competing services offer more generous access and faster availability, users eventually face a very simple decision:
Why should Grok remain their primary AI assistant?
We would rather not have to ask that question.
Fix the limits. Give subscribers transparency. Give every paying SuperGrok user equal access to the latest Grok models.
That is all we are asking.
Respectfully,
Dimitar Dimitrov
@thisisdimm
SuperGrok subscriber
