TL;DR
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
The Pragmatic Engineer’s report on 27 workplaces found wide variation in AI spending, from under $300 to more than $1,000 per engineer each month. CodeRabbit data showed PR volumes more than doubled among users over a year, but the report found that companies generally cannot say whether the extra output justifies the cost.
A report by The Pragmatic Engineer found that companies’ spending on AI coding tools varies widely, while available measures show higher pull request (PR) volumes among some users but do not establish whether the added output pays for the tools. The report gathered spending and engineering-leader accounts from 27 workplaces and analyzed aggregate data shared by CodeRabbit and Gitar.
CodeRabbit said its anonymized data from developers using its AI code review platform showed a 113% increase in weekly PRs and a 127% increase in monthly PRs. A separate comparison tracking the same group of 20,000 developers over a year showed a 128% increase. These are measures of PR activity, not a direct assessment of code quality, time saved, or financial return.
Gitar, an AI code verification agent, shared data from hundreds of customers showing different levels of PR output per developer. Its report categorizes some teams as shipping more than 60 PRs per developer per week. The Pragmatic Engineer’s author cautioned that team culture may influence these rates: engineers can adapt to the pace their workplace treats as normal.
Spending also differed sharply across the workplaces interviewed. The report describes a two-person .NET company spending $3,000 per developer monthly, while a publicly traded company with 1,500 engineers capped monthly spending at $150 per developer. Some companies reported that API costs had fallen by about half after summer peaks, while others anticipate higher bills as subsidized subscriptions end. The report says some firms’ costs could rise five to ten times when subsidies expire.
PR Volume Is Not a Return Measure
For engineering leaders, the findings show why tool budgets and output metrics are not enough to settle the question of ROI. A higher PR count may reflect more work being submitted, but the figures provided do not show whether the code is better, whether review and maintenance costs have changed, or whether teams are completing valuable work faster.
The costs are also difficult to compare. Some employees use subscriptions whose full price may be subsidized or obscured, while API-based use can fluctuate with model choice and usage. The report says spending caps are common among medium and higher spenders, though respondents described those limits as easy to avoid. Companies may need to track both actual costs and downstream outcomes before deciding whether broader AI access is worthwhile.
As an affiliate, we earn on qualifying purchases.
How Companies Track AI Use
The report began with a question from an engineer whose company was reviewing its AI budget: what do other businesses spend per engineer, and how has PR throughput changed? The author contacted engineering leaders and senior engineers at 27 workplaces, anonymizing most company names at their request, and supplemented those interviews with aggregate platform data.
PR throughput was chosen because companies can usually measure it and compare periods before and after adopting AI tools. But the report presents it as a readily available indicator, not a complete productivity measure. It also notes that spending patterns did not track company size in this sample. A trading firm in the report, for example, said its total AI spend reached $8 million a month in September, up from $500,000 in January; its cited average was $1,600 per developer monthly, though traders were described as major users too.
“Lots of traders are using Claude. Engineering is learning the importance of rate limiting for traders who accidentally run expensive queries without realizing it.”
— An engineer at the trading firm, quoted by The Pragmatic Engineer
As an affiliate, we earn on qualifying purchases.
Costs and Productivity Still Unmeasured
The report does not establish that AI tools caused the reported increase in PRs. The platform figures cover users of particular services, and the supplied material does not provide a controlled comparison that rules out other changes in staffing, work practices, or demand. Nor does it establish whether more PRs led to more production-ready features or better business results.
Company spending figures are self-reported accounts from a limited set of workplaces, with many names withheld. The report’s comparisons also cover different time windows and types of spending, including subscriptions and API use. It remains unclear how representative the 27 workplaces are, what share of each company’s total bill comes from engineers, and how often higher output offsets costs for review, rework, or infrastructure.
As an affiliate, we earn on qualifying purchases.
Companies Face Subscription Changes
Companies whose subsidized AI subscriptions are ending will need to decide whether to absorb higher bills, restrict access, or shift toward lower-cost models and infrastructure. The report says some businesses are already changing model choices or using open models as API costs fall from summer highs.
The next useful evidence will come from companies comparing total AI costs with measurable engineering and business outcomes over consistent periods. Until those figures are available, PR throughput can indicate a change in activity, but it cannot by itself answer whether the spending is worth it.
AI development productivity software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
How much are companies spending on AI per engineer?
The 27-workplace report found a wide range. Examples included a publicly traded company capping spending at $150 per engineer per month and a two-person startup spending $3,000 per developer monthly. These examples are not a representative market average.
Did AI use increase pull request volume?
CodeRabbit reported increases of 113% in weekly PRs and 127% in monthly PRs among users in its data. A tracked group of 20,000 developers showed a 128% increase over a year. The figures show an association in platform data, not proof that AI alone caused the change.
Does the report show that AI coding tools deliver positive ROI?
No. The report found higher PR volumes in some datasets and substantial variation in spending, but said few companies could answer the ROI question. PR counts do not measure code quality, time saved, or business value on their own.
Why might AI bills change?
Costs can vary with usage, model selection, API pricing, and subsidies. The report says some companies saw API spending roughly halve from summer peaks, while others expect costs to rise five to ten times when subsidized subscriptions end.
Source: rss
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
