AI coding costs may surpass developer salaries by 2028: Gartner
Synopsis
Gartner's latest report delivers a stark warning: AI coding tools could cost more than the developers they are meant to assist — and as early as 2028. The culprit is not the technology itself but the absence of governance, with token consumption spiralling unchecked as organisations rush from pilots to full-scale deployment.
Key Takeaways
Gartner Inc. warns that AI coding costs may surpass average developer salaries by 2028 .
Rising LLM token consumption and the shift to consumption-based pricing are the primary cost drivers.
Nitish Tyagi , Senior Principal Analyst at Gartner, flagged that developers prioritise speed over cost, making top-down governance essential.
Many AI coding vendors lack transparency in how token usage is calculated and billed , limiting enterprise cost forecasting.
Key risk factors include ungoverned agent autonomy , bloated context windows , and absent feedback mechanisms.
Gartner recommends a use-case-driven framework , model-task alignment, context engineering, and formal cost controls.
The cost of deploying AI coding agents could exceed the average developer salary by 2028, driven by surging large language model (LLM) token consumption and a broad industry shift toward consumption-based pricing, according to a new report by Gartner Inc. released on Wednesday, 24 June. The warning comes as organisations accelerate the move from pilot programmes to full-scale AI coding deployments — often without the financial controls to match.
The Token Cost Problem
AI tokens are the discrete units of data processed by generative AI models, and token consumption is the primary cost driver for AI coding tools — particularly under consumption-based pricing structures that are fast replacing traditional seat-based licences. As more developers use these tools at scale, token volumes compound rapidly, creating cost curves that many enterprise budgets have not accounted for.
Nitish Tyagi, Senior Principal Analyst at Gartner, said organisations are underestimating the financial exposure. 'Organisations are rapidly moving from experimentation to scaled deployment of AI coding agents, but many are underestimating the financial impact of rising token consumption,' he said.
Why Governance Is the Missing Link
Tyagi cautioned that cost discipline will not emerge organically from developer behaviour. 'Token discipline will not emerge through developer choice alone, as developers tend to optimise for speed and convenience over cost efficiency. Without a governed engineering operating model, costs can escalate faster than the productivity gains these tools are designed to deliver,' he added.
The report identified three structural drivers of overspending: ungoverned autonomy in agent-driven workflows, bloated context windows, and the absence of structured feedback mechanisms to optimise usage. Together, these factors create a spending environment where budgets are depleted faster than value is realised.
Transparency Gap Among Vendors
A significant concern flagged in the report is the lack of pricing transparency among AI coding agent vendors. Many providers do not clearly disclose how token consumption is calculated and billed, which limits enterprises' ability to forecast costs accurately or track cost-to-value outcomes. Without visibility into token usage across development tasks, organisations risk both budget overruns and an inability to justify AI spend to leadership.
'Software engineering leaders are increasingly concerned as token-driven AI spend becomes harder to justify, with budgets often being depleted earlier than expected,' Tyagi noted. He added that most organisations still lack the maturity and frameworks to effectively measure cost versus business impact.
What Organisations Should Do
The Gartner report outlined a set of recommended actions for software engineering leaders. These include establishing a use-case-driven decision framework, aligning model selection with task complexity, mandating context engineering practices, and implementing robust governance and cost controls. The underlying message is that AI coding tools require the same financial discipline as any enterprise infrastructure investment — something many technology teams have yet to build.
As AI coding adoption deepens across Indian and global enterprises, the cost question is likely to become a boardroom issue well before 2028.
Point of View
But the deeper issue is structural: enterprises adopted AI coding tools under seat-based pricing logic and are now being repriced into a consumption model they did not budget for. Token costs are invisible until they are not — and by then, the overspend has already happened. Indian IT services firms, which are deploying AI coding agents at scale to protect margins, face a particular irony: the tools meant to reduce headcount costs could, without governance, produce cost lines that rival the salaries they were meant to offset. The accountability gap — who owns token spend, engineering or finance — remains unresolved in most organisations.
NationPress
24 Jun 2026
Frequently Asked Questions
Why could AI coding costs exceed developer salaries by 2028?
According to a Gartner report, the shift from seat-based to consumption-based pricing for AI coding tools means costs scale directly with token usage. As organisations deploy AI coding agents at scale, token consumption rises sharply, and without governance controls, total spend could surpass the average developer salary by 2028.
What are AI tokens and why do they matter for cost?
AI tokens are the units of data processed by generative AI models. Under consumption-based pricing, every interaction with an AI coding tool consumes tokens, and the cost accumulates with each query, code generation request, or context window expansion — making token volume the central variable in AI tool expenditure.
What does Gartner recommend to control AI coding costs?
Gartner recommends that software engineering leaders establish a use-case-driven decision framework, align model selection with task complexity, mandate context engineering practices, and implement formal governance and cost controls to prevent budget overruns.
Why are AI coding vendors a concern in this context?
Many AI coding agent vendors do not clearly disclose how token consumption is calculated and billed. This lack of transparency makes it difficult for enterprises to forecast costs accurately or measure cost-to-value outcomes, increasing the risk of unplanned budget depletion.
Who is most affected by rising AI coding costs?
Organisations that have moved beyond experimentation into scaled deployment of AI coding agents are most at risk. Software engineering leaders across sectors — particularly those in IT-intensive industries — face growing pressure to justify token-driven AI spend as budgets are depleted faster than expected.
