Per-Token vs Per-Seat Pricing: How AI Vendors Will Restructure in 2026

The Pricing Restructuring

Through 2024-2025, almost every AI vendor charged per-token (raw API providers) or per-seat (SaaS-shaped products like Cursor, ChatGPT Plus). By 2026, per-token and per-seat are joined by per-task and per-outcome pricing. Vendors are restructuring as they figure out how to align price with value.

This piece walks through the four pricing structures and what each one optimizes for.

The Four Models

flowchart TB
    Models[2026 AI pricing] --> Token[Per-token]
    Models --> Seat[Per-seat]
    Models --> Task[Per-task / per-call / per-action]
    Models --> Out[Per-outcome]
    Token --> CommU[Use: API providers, infrastructure]
    Seat --> Comm2[Use: developer tools, productivity SaaS]
    Task --> Comm3[Use: agentic platforms, voice agents]
    Out --> Comm4[Use: results-driven SaaS, emerging]

Per-Token

The original. OpenAI, Anthropic, Google, and most API-shaped providers charge per million input and output tokens. Maps to underlying compute cost cleanly. Predictable for vendors; variable for buyers.

• Pro for vendor: aligns revenue with cost; scales with usage
• Pro for buyer: pay only for what you use
• Con for buyer: bills are unpredictable; very variable workloads create budget angst

Per-Seat

The SaaS standard. Cursor, Windsurf, GitHub Copilot, ChatGPT Plus all use per-user-per-month pricing. Predictable for both sides; does not scale linearly with usage.

• Pro for vendor: predictable revenue
• Pro for buyer: predictable budget; aligns with HR processes
• Con for vendor: heavy users subsidized by light users
• Con for buyer: pay for seats that may be lightly used

Per-Task

The 2025-2026 emergence. Pay per call, per ticket resolved, per agent action. Common for voice-agent platforms, many vertical agent products.

• Pro for buyer: aligns price with value (per resolved ticket vs per LLM call)
• Pro for vendor: scales with successful work
• Con for vendor: needs to define and measure tasks; metering complexity
• Con for buyer: still variable, just at a different unit

Per-Outcome

The most aspirational and least mature. Pay only when a defined outcome is achieved — a sale, a saved customer, a resolved case. Sierra (Bret Taylor) and several outcome-based AI vendors have leaned into this in 2026.

• Pro for buyer: pure value alignment
• Pro for vendor: can charge premium prices for proven outcomes
• Con for both: defining outcomes is hard; attribution disputes are common
• Con for vendor: revenue is delayed; cash flow strain at scale

Where Each Model Wins

flowchart TD
    Q1{Highly variable<br/>workloads?} -->|Yes| Token2[Per-token or per-task]
    Q1 -->|No| Q2{Productivity tool<br/>per-user usage?}
    Q2 -->|Yes| Seat2[Per-seat]
    Q2 -->|No| Q3{Defined outcome<br/>measurable?}
    Q3 -->|Yes| Out2[Per-outcome]
    Q3 -->|No| Task2[Per-task]

A Mixed Future

The 2026 reality is that no single pricing model wins. Most vendors offer some combination:

• API providers: per-token + volume tiers + commitment discounts
• Productivity tools: per-seat + usage caps with overage
• Agent platforms: per-task or hybrid per-seat + per-action
• Outcome-driven: per-outcome with floor / commitment

The pricing complexity has gone up, not down.

Buyer Considerations

For procurement teams in 2026:

• Compare apples-to-apples on TCO, not unit pricing
• Watch for hidden minimums, overages, "fair use" caps
• Negotiate enterprise discounts based on total commitment
• Build in price-protection clauses for multi-year deals
• Understand the vendor's pricing direction; sticker price today may not last

Vendor Considerations

For AI vendors deciding pricing:

• Match pricing to your unit economics
• Consider how your pricing affects buyer decisions (per-seat may discourage adoption)
• Per-outcome pricing requires deep alignment on measurement
• Hybrid models (small base + usage) often work better than either extreme
• Pricing changes are expensive (customer trust); design for stability

What's Coming in 2026-2027

• More verticalized pricing (per-call for voice; per-PR-merged for code; per-prediction for ML platforms)
• More outcome experimentation, with mixed results
• Pricing transparency tools (cost calculators, public pricing pages becoming norm even for enterprise products)
• Dynamic pricing experiments (peak-vs-off-peak, regional)

What This Means for Buyers in 2026

Three rules of thumb:

• Avoid pricing models with unbounded variability unless you have a strong cost-control story
• Per-seat models for productivity tools where usage is roughly proportional to headcount
• Per-task or per-outcome models for production AI where results are measurable
• Always model worst-case usage, not average

Sources

• "AI pricing models" a16z — https://a16z.com
• "Outcome-based pricing" SaaSPath — https://www.saaspath.io
• Sierra "outcome-based AI" — https://sierra.ai
• "Software pricing" OpenView Partners — https://openviewpartners.com
• Pricing strategy research — https://www.priceintelligently.com