HaiAI123

Curated Global AI Tools Directory

Industry update
Flat

Parcha's agent PRs push Greptile review bill from $180 to over $7,000

11/21/2026 — 11/21, 17:17·1 sources (all)·1 reports (all)

Key takeaways

  • According to Stack Futures, Parcha's coding agents filed so many pull requests that Greptile's per-review pricing pushed its code review bil…

Background & analysis

According to an account from Stack Futures, Parcha's coding agents began submitting large volumes of pull requests, which pushed up its code review costs on Greptile. The trigger was a mundane change in Greptile's pricing: the service moved from a flat per-seat model to charging $1 per review beyond the first 50. Under the new scheme, Parcha's monthly review bill climbed from $180 in December to $6,451 in October, and reached $7,018 so far in November, roughly 36x the original level.

To address the rising bill, Parcha opted to use a low-cost model and train its own reviewer on proprietary pull request data, bringing the cost of a single review down to about $0.18. According to the same account, this in-house reviewer detected 42% of real bugs on 30 held-out pull requests in its repository, compared with a 26% detection rate for Greptile. The available material only documents this approach and does not describe any developments that followed.

AI-generated from 1 reports · updated 17 hours ago

updream
24-hour heatHeat index 60 · peak 97 · 16h ago
24 hours agonow

Reports on this story headlines open the original

Yesterday
  1. According to Stack Futures, Parcha's coding agents opened so many pull requests that Greptile's per-review pricing pushed its monthly bill from $180 to $6,451 in October and $7,018 so far in November. Rather than switching models, Parcha trained its own reviewer on a cheap model using its own PR data, cutting a review to about 18 cents and catching 42% of real bugs on 30 held-out PRs versus Greptile's 26%.

    DEV Community · AIAI score 57

Other stories people are talking about

How is heat calculated?About the method

Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.

This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.

Surge
Discussion rising fast
New
First report within 6 hours
Rising
Still gathering discussion

Back to AI industry updates →