When a Single Query Costs $5,000: A Devs’ Nightmare with Google Cloud’s BigQuery

Brett Tabkeon

Multiple Solana users reported receiving multi‑thousand‑dollar bills after executing just one - or a few - queries via Google Cloud’s BigQuery. One anonymous dev shared that their billing, typically in the “few hundred dollars” per month range, shot up to a staggering $18,000—all triggered by just three queries, each initially billed at over $5,000 before being adjusted by support to $4,000 each

Predatory Pricing or Poor Defaults?

Voices across the crypto space criticized BigQuery’s lack of spending safeguards:

“They intentionally don’t let you set hard stops,” remarked Ermin Nurovic, co‑founder of the Flat Money synthetic dollar protocol, adding, “Your Google Cloud function got stuck in a recursive loop costing you thousands? Too bad.”

Developers argued this pay‑as‑you‑go model, especially without partition checks or query limits, can turn simple mistakes into financial catastrophes.

More Stories of Pain

Another pseudonymous developer recounted an “accidental” query that scanned terabytes of Solana data—resulting in a $5,000 bill. Luckily, because their company had local contacts at Google, they managed to escalate the issue and secure a refund. Post‑incident, they vowed never to query blockchain data without first verifying partitions .

The Root of the Problem: Lack of Budget Controls

BigQuery is a powerful, serverless data warehouse capable of handling massive datasets using SQL and built‑in AI tools. However, unlike many cloud services that allow users to set spending caps, BigQuery doesn’t offer hard stops—meaning runaway scripts or unintended full‑table scans can trigger enormous unanticipated charges.

Solana’s integration with BigQuery (launched in October 2023) enables developers to access rich historical blockchain data like whale transactions, NFT sales, and more. While incredibly useful, it also puts them at risk of massive compute costs—especially if queries aren’t tightly scoped.