Source-linked reports: Amounts and outcomes come from the original reports and linked replies.
An idle container kept running and cost $78.74.
Community user sdultsin says a Cloudflare Container left running idle for about 26 days was charged $78.74 although the invoice showed zero Workers, Durable Objects and R2 usage; the charge was memory ($72.46), disk ($3.35) and vCPU ($2.93) allocation. They deleted the container, requested a refund through a support ticket and say there was no response after about 10 days.
What happened
- Reported cause
- Active duration
- Incident date
- Date not recorded
- First reported
- 2026-03-09
- Billing outcome
- Request submittedAuthor-reported
- Case progress
- Not established
- Public Cloudflare reply
- Not linked
Latest reported outcome
The author says a support ticket requesting a refund had no response after about 10 days.
View this update in the timelineOriginal sources
Each entry links to the report or reply it describes.
Case timeline
Author requests a refund
Open original sourceThe author reports $78.74 for an idle container and requests a refund.
More to read
A budget alert is a horn, not a brake
@GUYUE_LGY points to Cloudflare’s Budget alerts documentation: alerts send email when spend crosses a threshold but do not pause or cap usage. Before letting AI run unattended, the author checks what actually happens at the line.
forsalebypwner: an exorbitant Workers Containers charge with no usage visibility.
In a Hacker News comment, forsalebypwner says that one month they were charged exorbitantly for Workers Container usage and that there was no way to monitor or track usage until the invoice arrived.
Removing unneeded Durable Objects from MCP servers on Workers
Justin (@interjc) describes moving read-only MCP servers from McpAgent, which starts a Durable Object per session, to stateless createMcpHandler, and shares a prompt for auditing Workers bindings and runaway usage patterns.