
Tech • AI • Robotics • Game
Grokbot users are reporting faster depletion of weekly quotas after Elon Musk said the service had become too expensive to operate relative to subscription revenue, raising concerns over reduced value and unclear limits.
On November 1, Elon Musk responded to a request for a usage reset by saying the cost of serving Grokbot was “unsustainably high” compared with revenue. The remark amounted to a public acknowledgment that operating expenses had outpaced what subscriptions were bringing in. No official figures were released on losses, quota sizes, or the scale of any changes.
Around the same time, users began reporting that their weekly allowances were running out days earlier than expected. Some said their usage meters jumped from roughly half consumed to fully used over a weekend without corresponding activity. Others described the change as an effective reduction in service despite no announced price cut or plan change.
The extent of the problem remains unclear because some discussion has pointed to a billing or dashboard bug. That leaves open the possibility that at least part of the spike in reported usage reflects inaccurate metering rather than a real quota cut. Even so, Musk’s statement on costs has reinforced the view that some tightening may be deliberate.
The controversy comes as Grokbot has been expanding quickly. Daily users were described as rising from under 100,000 in September to more than 600,000 by late October, with paid weekly users projected around 800,000. That pace was framed internally as one of the fastest product ramps in Musk’s businesses, but it also implies sharply rising compute bills.
One likely cost factor is the use of multiple high-end models for different tasks. In October, Musk said Grokbot would use the best model available for a given request, including Claude Opus 5.5. If a significant share of user tasks is being routed to premium third-party models, inference costs could rise well above those of in-house systems.
Generous background usage had been a major reason many users chose Grokbot over rivals such as Claude, ChatGPT, and Meta-linked assistants. The appeal was that users could hand over multiple tasks without constantly watching a quota meter. If those limits have tightened while subscription pricing remains unchanged, the product’s competitive advantage may narrow quickly.
Users on X can purchase additional usage credits after reaching their limit. However, reports indicate there is no default monthly spending cap on that add-on system. That creates a risk of unexpectedly high charges unless users manually set a limit.
Several central questions remain unanswered: whether the change is temporary, how much weekly allowances were reduced, whether a reset will be offered, and how much of the issue stems from a dashboard error. The company also has not publicly disclosed the exact weekly quota, making it difficult for subscribers to measure any reduction precisely.
Similar complaints surfaced in May over caps on image and video generation, after which limits were later increased. That history has led some users to expect a partial reversal if more computing capacity comes online. For now, though, frustration appears tied less to the existence of limits than to the lack of transparency around them.
The dispute over Grokbot highlights the tension between rapid AI growth and the heavy cost of serving advanced models at scale. Without clear disclosure on quotas, billing accuracy, and future policy, user trust may become as important as compute capacity.
Ask a question