Hi everyone,
I’ve been following the recent discussions about Bugbot’s transition from the $40 flat rate to usage-based pricing. While I understand that low-volume users (doing maybe 8-10 PRs a month) might favor this, this update completely breaks the tool for power users who deeply integrate Bugbot into their daily workflow.
Here is the reality of using Bugbot for heavy, continuous development—such as deep, complex refactorings of legacy CMS architectures to modern PHP standards:
1. A Pull Request is not a single action
For an iterative workflow, the ideal trigger mode is “Every Push”. This means a single PR can easily trigger 7 to 10 Bugbot runs as the code is continuously refined, debugged, and updated before the final merge. Under the old $40 flat rate, this was exactly what made Bugbot so incredibly valuable: it acted as a continuous, asynchronous pair programmer.
2. The new math is completely unsustainable
At the newly confirmed $1.00 - $1.50 per run (not per PR), a standard month of 20 PRs (averaging 7 runs each) will now cost roughly $150 to $200+ on top of the Ultra/Pro subscription. That is a massive, unjustifiable price hike. This pricing model actively penalizes continuous AI review and forces developers to use Bugbot merely as a final, static linter right before merging just to avoid exploding costs. That defeats the entire purpose of an AI coding assistant.
3. Misleading communication regarding “Effort Levels”
The official email announcement explicitly promoted that users can now “determine the effort Bugbot applies” to manage costs or review depth. However, as confirmed by staff in other threads, this feature seems to be locked exclusively behind Team plans. Promoting a cost-management feature to individual Pro/Ultra users who won’t even have access to it is incredibly frustrating and feels like a bait-and-switch.
4. A fatal flaw for Agentic Workflows and “Vibe Coding”
Cursor positions itself as the IDE for the future of AI and autonomous agents. However, for developers utilizing orchestrator workflows, subagents, or vibe coding—where code is iteratively generated, pushed, and tested by AI—a pay-per-run model is completely unmanageable. An autonomous loop triggering “Every Push” could silently drain hundreds of dollars. This pricing effectively kills the ability to securely use Bugbot alongside automated agentic workflows.
5. Cannibalizing the Ultra Plan allowance
As a power user, my data allowances and fast requests on the Ultra plan are already heavily utilized just by coding. If Bugbot now aggressively eats into the included data allowance—or forces massive on-demand charges on top—it effectively forces developers to choose between writing code and reviewing it. If Bugbot drains my entire budget halfway through the month, what is the point of the Ultra plan? The tool meant to review code is stealing the resources needed to write it.
By catering to users who barely utilize the tool, Cursor is essentially pricing out the power users who rely on it the most. It feels like this transition was not thought through at all.
Are there any plans to introduce a true power-user flat rate? Alternatively, will you allow current subscribers to upgrade to an annual plan before the June 8th deadline to lock in the old $40 rate for the upcoming year? Because right now, the financial punishment for writing clean, iterative commits is simply too high.