How AI Companies Are Figuring Out Pricing Beyond SaaS

AI pricing is in an experimental phase. Kate McAndrew, a venture capitalist, breaks down the four emerging pricing models for AI products and why SaaS subscriptions alone do not work for this category. She credits Meta Agrawal from Defy VC for articulating this framework clearly.

Why SaaS Pricing Does Not Fit AI

AI tools pull from traditional labor budgets in ways that are fundamentally different from SaaS software. When your product replaces hours of human work, a flat monthly subscription undersells the value. AI requires more dynamic pricing that reflects the actual impact it delivers.

Kate McAndrew speaking to camera with the text "Venture Capitalist's perspective on the Wild Wild West of AI"

The Four Pricing Models

Fixed Cost (SaaS): A flat subscription fee. This still works for everyday use cases where the value is predictable and consistent.

Input-Based: Charging by consumption. This means pricing based on the amount of data processed or the number of queries made. The more you use, the more you pay.

Outcome-Based: Pricing tied to the actual outcome the tool delivers. For example, an AI that automates administrative tasks in healthcare would charge based on tasks completed, not queries run.

Hybrid: Some combination of fixed cost, input-based, and outcome-based pricing. Most companies will likely land here, mixing two or three of these models together.

A LinkedIn post from Meta Agrawal at Defy VC outlining the four AI pricing models, visible behind Kate as she explains the framework

How Companies Are Choosing

Right now, there is no consensus. Companies are running experiments with their early B2B customers (design partners) to figure out what budgets they can pull from and what pricing model those customers will accept. It is genuinely the wild west, which Kate finds exciting rather than alarming.

Key Takeaways

AI products need pricing models beyond flat SaaS subscriptions because they replace labor, not just software
The four emerging models are fixed cost, input-based, outcome-based, and hybrid
Most AI companies are still experimenting with design partners to figure out what works

Resources

Defy VC -- Venture capital firm; Meta Agrawal's posts on AI pricing are worth following