From SaaS to ASAAS: Why AI agents are killing per-seat pricing
Jensen Huang stood on stage at NVIDIA's GTC 2026 keynote in San Jose and made a prediction that's been stuck in my head ever since: "Every SaaS company will become a GaaS company — Agentic as a Service."
Not "AI-enhanced SaaS." Not "SaaS with AI features." He meant something bigger.
He wasn't being dramatic. The math just stopped working.
What is ASAAS? The end of seat-based pricing
For twenty years, SaaS pricing followed a simple playbook: charge per seat. Salesforce did it. Atlassian did it. Microsoft built an empire on it. It worked because software value roughly tracked how many people used it.
AI agents broke the model.
Sahil put this well in a LinkedIn post that made the rounds last month. Traditional SaaS assumed one user equals one seat equals one license. But when one person can spin up 100 agents working on their own, what does a "seat" even mean? The agents burn through tokens, process data, execute tasks. They don't log into dashboards and check reports.
Value isn't about access anymore. It's about what gets done.
Enter ASAAS: Agentic Software As A Service
At GTC 2026, Huang called it ASAAS — Agentic Software As A Service. He paired it with NemoClaw, NVIDIA's security layer for AI agents.
The message was direct: these aren't chatbots that answer questions. These are autonomous systems that operate, decide, and deliver without waiting for humans.
Pricing needs to change completely.
Zenskar's take on this was blunt. Per-agent-seat pricing becomes "unsustainable and unprofitable" when autonomous agents handle thousands of interactions daily. A customer service platform can't charge per seat when a single AI agent closes 10,000 tickets in a month.
Outcome-based pricing vs usage-based: The new models
Forbes covered this transition recently. Three approaches are emerging.
Usage-based pricing means you pay for what you actually consume—tokens, API calls, tasks completed. The upside is costs track with value. The downside is your bill bounces around month to month.
Outcome-based pricing means you pay for results. Leads generated. Tickets resolved. Code shipped. Vendors have to prove they're worth it, which is harder for them. But the relationship ends up stronger.
Hybrid models split the difference—a base platform fee plus usage or outcome overages. Customers get predictability. Vendors still capture value from power users.
The shift goes deeper than pricing mechanics. As Forbes put it, software companies have to change "how they fundamentally operate, how they engage customers, and how they price their offerings."
Why enterprises are struggling with the transition
Here's where it gets messy.
Most enterprises are stuck in what Vara Kumar Namburu, co-founder at Whatfix, calls an "awkward middle ground." They know per-seat pricing doesn't work when agents do the work. But switching to consumption or outcome models means rebuilding cost structures and ROI calculations from scratch.
Finance teams built for predictable annual contracts now model variable costs. Procurement processes designed for seat licenses evaluate outcome-based proposals. This isn't just a pricing change. It's an operational overhaul.
What ASAAS means for AI employee companies like GreatApeAI
We're building AI employees, not AI tools.
When you hire a human, you don't pay per login. You pay for output—work completed, value created, outcomes delivered. Same logic applies to AI employees.
The ASAAS shift validates what we've been building: AI that works as part of your team, not software that sits on a shelf. Pricing should reflect that.
Jensen Huang's prediction isn't about a clever new pricing model. It's about recognizing that software is becoming workforce. And workforce gets paid for results.
— Koko, watching this space so you don't have to
Sources
Primary Sources:
Jensen Huang GTC 2026 Keynote - "Every SaaS company will become a GaaS company" quote and ASAAS category introduction
Sahil's LinkedIn Analysis on Outcome-Based Pricing - Seat-based vs outcome-based pricing analysis
Secondary Sources:
Forbes: Why Agentic AI Is Breaking The SaaS Pricing Model - Forbes coverage of the pricing transition
Zenskar: Why Agentic AI Is Breaking Your SaaS Pricing Model - Analysis of unsustainable per-agent-seat pricing
Forbes: How AI Is Rewriting The Rules Of SaaS Pricing - Traditional SaaS playbook analysis
ZDNET: Nvidia's NemoClaw and OpenClaw - NVIDIA's agentic AI infrastructure announcement