A growing shift is materializing in the IT services market as software vendors move away from traditional fixed software pricing and towards models based on AI token consumption, effectively redefining the sources of margin and exposing providers to variable cost structures. This transition is embodied by companies such as Flamingo, which has constructed an MSP software suite and prices access at $1 per device per month, including a token allotment, but whose actual profits—and partner costs—are derived from overage on AI token consumption. The fundamental mechanism is that margin is increasingly being earned from metered AI usage rather than the flat software license model that previously dominated the sector.
The most consequential development highlighted involves Flamingo’s approach: software fees serve primarily as an entry point, while ongoing, uncapped AI token consumption drives real costs for providers. According to Flamingo, most partners use 10 to 20 times their monthly token allocation, with the overage billed at rates that track costs from AI lab providers such as Anthropic and OpenAI. These rates are variable, not transparently published, and fluctuate as AI model costs change. Flamingo claims incremental margin via internal consumption efficiency, particularly as it plans to host its own models and increase token allotments without reducing partner pricing, consolidating margin from lowered costs.
Additional evidence is provided by Integris, a private equity-backed MSP that launched a bundled AI service, Core, currently priced per seat or device despite the product’s purpose being automation and labor reduction. Integris representatives indicated that no stable, customer-aligned outcome-based pricing model has emerged, a challenge echoed by other large MSPs and exemplified by Salesforce’s challenges in usage and outcome-based AI billing. At every level—from the vendor building software on top of AI labs to the MSP implementing services for clients—pricing mechanisms remain unsettled, with risk of misalignment and unpredictability being pushed downstream.
For MSPs and technology leaders, these developments introduce new operational risks, chiefly around cost forecasting, pricing transparency, and margin management. The shift to consumption-based and hybrid models increases the burden of monitoring both cost and value delivered from AI-powered services, while lack of clear, measurable outcome units limits the feasibility of outcome-based billing. Providers may find themselves forced to absorb cost variability while delivering fixed-fee services, or to renegotiate contracts and client expectations as variable pricing becomes standard. As AI-related spending rises and billing models remain opaque, effective governance, cost auditing, and risk management become central to sustainable operations.