Market data and pricing benchmarks revealing how AI agents are transforming billing models, cost structures, and revenue design for software companies
AI agents are moving from experimentation toward sustained enterprise investment, and pricing them has become one of the most complex challenges in modern software monetization. Unlike traditional SaaS products with predictable seat-based models, AI agents consume variable compute resources, process unpredictable token volumes, and increasingly bill against outcomes that resist simple per-unit measurement. Companies building AI-powered products need usage-based billing infrastructure capable of handling this complexity while maintaining the financial accuracy that enterprise customers demand.
Key takeaways
- Hybrid pricing is now the default, not the exception: 95% of the 80 AI agent companies in Orb's 2026 study use hybrid pricing, up from 92.4% a year earlier, and Orb describes hybrid pricing as the baseline
- Usage-based pricing is near-universal: 91.3% of AI agent companies meter and charge for usage, while 71.3% still carry a subscription component
- Outcome-based pricing remains rare and slipped year over year in Orb's sample: 3.8% of the sample uses outcome-based pricing, down slightly from 4.5% in 2025, even as vendors publish headline prices tied to completed customer-support interactions
- Published customer-support resolution rates vary by plan and outcome: Gorgias lists $0.90 per resolved interaction on most plans, while Starter plans begin at $1, and Intercom lists $0.99 for a resolution outcome, while the cited Zendesk automated-resolution tier documentation does not state a list rate
- Enterprise budget mix could shift materially: Deloitte, citing Gartner, projects that by 2030 at least 40% of enterprise SaaS spend could move toward usage-based, agent-based, or outcome-based models
- Pricing remains in motion: 37% of companies plan to change their AI pricing model in the next 12 months, reinforcing the need for pricing agility
How AI agent companies actually price today
The clearest picture of current practice comes from Orb's 2026 study of 80 AI agent companies, supplemented by Deloitte's 2026 SaaS predictions, McKinsey's 2025 pricing-meter analysis, and ICONIQ's 2026 AI snapshot. Orb analyzed 80 AI agent companies at a 90% confidence level. Orb's AI-agent sample shows hybrid monetization as the baseline, while the broader sources document continued experimentation across pricing components and meters.
1. 95% of AI agent companies use hybrid pricing
Orb's 2026 research found that 95% of the 80 AI agent companies studied combine multiple pricing mechanisms rather than relying on a single model. Orb describes hybrid pricing as the baseline, making the ability to operate layered pricing without an engineering burden a core execution requirement.
2. 91.3% of AI agent companies use usage-based pricing
Usage-based pricing appears in 91.3% of the companies in Orb's 2026 sample, up from 83.3% in 2025. Orb describes usage-based pricing as the second most dominant component after hybrid pricing, increasing the importance of accurate, real-time usage metering.
3. 71.3% of AI agent companies still charge a subscription
Subscriptions have not disappeared. 71.3% of the companies studied retain a subscription element, up from 62.1% in 2025. Orb describes subscription pricing as a flat recurring fee for access and notes that subscriptions create a predictable revenue floor.
4. 94.7% of subscription-charging companies also charge for usage
Among companies that use subscriptions, 94.7% pair subscriptions with usage. Subscription-only pricing is uncommon among companies that retain a subscription component.
5. Outcome-based pricing is rare and slipped year over year
Despite the volume of commentary, outcome-based pricing sits at just 3.8% in Orb's 2026 sample, down slightly from 4.5% in 2025. Orb attributes the gap to instrumentation rather than appetite: defining, measuring, and attributing outcomes reliably is hard, and most billing infrastructure was not built to support it.
6. Hybrid adoption increased modestly from 92.4% to 95%
Within Orb's own year-over-year series, hybrid pricing moved from 92.4% in 2025 to 95% in 2026. Orb says the data moved only modestly and describes hybrid pricing as the baseline.
7. 40% of AI-native pricing meters are activity-consumption based, versus 17% among incumbents
McKinsey's 2025 pricing-meter analysis found that 40% of meters used by AI-native companies are activity-consumption based, versus 17% among incumbents. AI natives also use 35% flat-fee, 10% successful-outcome, and 15% capacity meters, while incumbents use 68% flat-fee, 2% successful-outcome, and 13% capacity meters. The distribution shows AI-native software leaning more heavily toward activity-based consumption while incumbents remain concentrated in flat-fee models.
8. At least 40% of enterprise SaaS spend could shift to usage, agent, or outcome models by 2030
Deloitte, citing Gartner, projects that by 2030 at least 40% of enterprise SaaS spending could move toward usage-based, agent-based, or outcome-based models. For revenue design teams, this projection signals that a large share of enterprise software budget may move toward variable or outcome-linked models.
Billable units and pricing metrics
The hardest question in AI agent pricing is not how much to charge but what to charge for. AI products can use a wider range of billable units than conventional seat-based SaaS pricing.
9. AI usage is billed across actions, compute, API calls, tokens, duration, or combinations
Deloitte describes usage-based AI agent pricing across actions and tasks, compute, API calls, tokens, duration, or combinations of those measures. These measures can also be combined within a pricing model.
10. Agent usage maps naturally to discrete events
Orb's 2025 analysis of AI agent pricing notes that agent activity is naturally represented by events such as calls, tokens, jobs, documents, or leads. Choosing the event is the pricing decision; everything downstream, from rate cards to invoices, inherits that choice.
11. Effort-based pricing is emerging as a distinct model
Orb's 2026 research identifies effort-based pricing, where compute time, number of steps, or task complexity become the value and usage measure. This sits between raw consumption and final outcomes, and it fits agentic workloads where a single task may trigger long, variable execution chains.
12. Gorgias's AI Agent bills by resolved interaction, not seats, messages, or tokens
Gorgias states that its AI Agent offering is not billed by seats, messages, or tokens, but by fully resolved interactions. This aligns the billable unit with completed customer interactions rather than underlying resource consumption.
13. Credit systems translate heterogeneous actions into a single unit
Orb supports prepaid credits that let customers buy a balance up front and draw it down as usage occurs. For AI products, one credit can abstract tokens, agent runs, and outcomes into a single unit, while enterprise customers can commit to prepaid balances over a quarter or year and draw them down with usage.
14. Flat-fee per-agent or digital-worker pricing remains a documented model
Deloitte documents flat fee and agent pricing, including a digital-worker salary-like model, as one of several emerging approaches. It remains one documented approach alongside usage-based and outcome-based pricing.
Outcome pricing benchmarks from first-party vendor documentation
Where outcome pricing does exist, the published mechanics are more precise than the summary coverage suggests. These figures come from vendor documentation rather than secondary aggregation.
15. Intercom charges $0.99 for a resolution outcome
Intercom's Fin documentation prices a resolution at $0.99, alongside procedure handoffs and disqualifications, for Fin over chat and email. Pricing is outcome-based, with at most one outcome charge per conversation.
16. Different outcomes carry different unit values
Intercom does not price all outcomes identically. Its documentation lists $0.99 for several outcome types and $9.99 for qualification, a roughly tenfold spread between the listed rates. For Intercom, outcome pricing is therefore a rate card rather than a single number.
17. Intercom caps billable outcomes at one per conversation
Intercom charges at most one outcome per conversation. Intercom states that this cap applies regardless of how many questions the customer asks or how many Procedures Fin runs in that conversation.
18. No charge when an eligible outcome is not achieved
Intercom does not charge when an eligible outcome is not achieved. This makes precise success definitions and auditability important to the pricing model.
19. Gorgias lists $0.90 per resolved interaction, with Starter beginning at $1
Gorgias prices its AI Agent at $0.90 per resolved interaction on most plans, while Starter plans begin at $1 per resolved conversation. A billable interaction is counted when the AI resolves a customer conversation entirely on its own.
20. Zendesk restructured packaging around automated resolution tiers
Zendesk introduced automated resolution tiers on May 18, 2026. Automated resolutions are the unit used to calculate and bill AI agent usage, accounts are credited with a resolution allowance, and the tiers are designed so customers do not overpay for straightforward tasks.
21. Only verified resolutions draw down the allowance
Under Zendesk's tiers, an assisted escalation and a contained resolution do not count against the resolution allowance. Verification runs when the conversation ends, and the end window depends on channel: email ends 72 hours after the first email, messaging ends 2 hours after the first message by default and can be increased to 72 hours, and voice ends immediately upon hangup. A large language model then evaluates the conversation and classifies successful cases as verified resolutions.
22. AI monetization increasingly spans multiple pricing rails
ICONIQ's 2026 snapshot, drawing on survey data from approximately 300 executives building AI products, found that 58% of companies include a subscription or platform pricing component, 35% use consumption-based pricing, and 18% use outcome-based pricing. The mix shows that multiple pricing components remain in active use.
Unit economics pressure on AI pricing
Outcome and resolution pricing depends on the relationship between unit price and delivery cost. Gartner's projection suggests delivery costs may remain material even as automation matures.
23. Gartner projects GenAI customer service cost per resolution will exceed $3 by 2030
Gartner predicts that generative AI cost per resolution for customer service will exceed offshore human agent costs by 2030, with cost per resolution exceeding $3, higher than many B2C offshore human agents. The forecast raises a unit-economics question for outcome-priced products, although vendor prices and Gartner's projected delivery costs are not directly comparable across products or time periods.
24. Gartner expects customer service AI goals to move beyond cost savings
Gartner says returns on AI customer service investments are far from guaranteed and that full automation will be prohibitively expensive for most organizations. Gartner expects leading organizations to use AI to drive customer engagement rather than to cut costs, and it projects that AI-related regulatory changes will increase assisted-service volume 30% by 2028. Together, those forecasts put pressure on a cost-savings-only rationale.
Pricing iteration as an operating discipline
The evidence indicates that AI pricing frequently needs reassessment as customer behavior and product economics become clearer.
25. 37% of companies plan to change their AI pricing model within 12 months
ICONIQ found that 37% of companies plan to change their AI pricing model in the next 12 months. Repricing is therefore a material near-term operating concern, making the operational cost and latency of pricing changes more important.
26. Pricing changes are driven by customer demand, competition, and margins
ICONIQ's 2026 snapshot says planned pricing changes are driven primarily by customer demand, competitive pressure, and margin concerns. Among companies experimenting with outcome-based pricing, 36% most often tie outcomes to cost savings and 18% to revenue generated, illustrating how pricing iteration follows value definitions as well as cost.
27. Pricing is becoming a continuously iterated system, not a one-time packaging decision
Orb's 2026 research frames AI agent pricing as an ongoing system rather than a single packaging exercise. Orb says companies that can continuously iterate, use real data, run safe experiments, and adapt quickly can create a meaningful advantage.
28. Execution is constrained by technical complexity, operational friction, and pricing knowledge gaps
Orb's 2026 study identifies technical complexity, operational friction, and pricing knowledge gaps as the practical limits on pricing execution. Orb describes execution, rather than uncertainty about pricing direction, as the real barrier.
29. Pricing is simulated against real usage before rollout
Orb supports pricing simulations against real product usage data before rollout. Modeling a proposed rate card against historical usage makes customer and revenue impact visible before a pricing change goes live.
Commercial guardrails and contract mechanics
Variable consumption creates variable bills, making budget predictability and commercial safeguards important to enterprise buyers. Guardrails have become part of the pricing model itself.
30. Caps, pooling, commitments, credits, and overage terms manage variable spend
Orb identifies caps, pooling across teams, seasonal overage protections, and credit bundles that roll over as mechanisms for managing consumption pricing. Orb also supports contract commits that can make committed revenue more predictable. These commercial terms have billing consequences, and each adds rules the billing system must enforce.
31. Hybrid structures pair a base fee with variable usage
A common hybrid structure combines a base platform fee with variable usage charges. The base or subscription component provides a more predictable revenue floor, while the variable component captures usage-driven upside.
32. Billable outcomes and attribution rules are defined to be auditable by the customer
Orb stresses defining billable outcomes and attribution rules precisely enough for customer-side reconciliation. When the invoice line is an outcome rather than a unit of input, the customer's ability to reconstruct the charge from event history becomes a commercial requirement.
33. Prepaid credits formalize committed consumption
Prepaid credits let enterprise customers commit to a balance over a quarter or year and draw it down as usage occurs. Orb supports traceable drawdowns, expiration handling, and top-ups so committed consumption can remain aligned with billing and finance workflows.
34. Real-time spend monitoring and threshold alerts improve spend control
Orb spend controls provide real-time spend monitoring, alerts, threshold-based notifications, and workflow triggers that help customers avoid unexpected charges and act when usage approaches defined limits.
35. Threshold billing creates threshold-triggered invoicing events
Threshold billing invoices when accrued usage crosses a defined dollar amount rather than waiting for a calendar boundary. For high-velocity AI consumption, this can reduce exposure to unpaid usage by billing before month end.
36. Dimensional pricing varies rates by region, instance type, and environment
Orb's dimensional price groups support pricing across multiple usage dimensions, such as region, instance type, and environment, using a single pricing configuration for dimension combinations. This lets a vendor vary rates across materially different usage contexts without maintaining a separate product for each combination.
Metering, revenue recognition, and finance operations
Pricing strategy is only as good as the system that executes it. In AI agent monetization, the metering and finance layer has become part of the pricing model rather than a downstream reporting concern.
37. Raw usage events underpin outcome attribution
A preserved layer of raw usage events gives vendors granular event history, while explicit attribution rules and audit trails connect that history to outcome billing. Together they support auditability, corrections, finance workflows, and reconstruction of why a specific outcome was billed.
38. Backdating and backfill make retroactive corrections possible
Orb documents backdating price changes and backfilling missing events, with invoices and credit ledgers recalculating automatically rather than through manual correction events. Orb's backdating and backfill capabilities are designed for scenarios such as late contract changes, outages, and missing events.
39. Metrics defined in SQL and versioned pricing decouple rate changes from code releases
Defining usage metrics in SQL over raw usage events allows teams to revise billable metric definitions using already captured event properties without introducing new instrumentation for those changes. Orb also supports version-controlled price changes with documented history, so repricing can be managed as an operational workflow instead of a fresh instrumentation project.
40. ERP-aligned finance workflows connect billing data to accounting systems
Orb's finance workflows include ASC 606-aligned reporting, enterprise contract tooling, and integrations with ERP and accounting systems. Its NetSuite integration pushes finalized records from Orb into NetSuite, helping keep invoicing and downstream accounting aligned.
41. Outcome-based pricing creates new revenue recognition judgments
Deloitte's June 2026 analysis notes that outcome-based pricing introduces new ASC 606 judgments for SaaS companies, including whether a service is a stand-ready obligation or a promise to deliver specified successful outcomes, along with the treatment of variable consideration. Outcome-pricing design can therefore affect revenue-recognition analysis as well as commercial design.
42. Auditable pricing changes reduce engineering dependence in revenue design
Orb combines version-controlled price changes with pricing simulations, giving finance and pricing teams a controlled path to test and operationalize changes while reducing dependence on deployment cycles. As pricing shifts from a static decision to a continuously iterated system, that operational agility becomes a durable advantage.
Strategic implications for billing and revenue design
These trends converge on a small number of concrete requirements for companies building AI-powered products.
Hybrid pricing demands flexible infrastructure
With 95% of companies using hybrid pricing in Orb's 2026 study and 94.7% pairing subscriptions with usage, billing systems must support layered structures natively:
- Base platform fees combined with per-event, per-action, or per-outcome charges
- Volume tiers with distinct unit economics above included allowances
- Dimensional pricing across regions, instance types, and environments
- Spend controls and threshold alerts so customers can manage variable cost in real time
Outcome-based billing requires event-level tracking
Pricing per resolution, per resolved interaction, or per task depends on granular event history. Event-level records can support:
- Accurate outcome attribution and the rules that govern it, including per-conversation caps where the pricing model uses them
- Retroactive adjustments and corrections through backdating and backfill
- Audit trails that enterprise customers can reconcile against their own systems
- Finance workflows and ASC 606-aligned reporting
Enterprise adoption accelerates billing complexity
As AI agents move from departmental tools to enterprise infrastructure, billing requirements expand:
- Dimensional pricing across regions, instance types, and environments
- Contract-based commitments with usage overages, pooling, and caps
- Credit systems for prepaid consumption, including drawdown and expiry handling
- Integration with ERP and financial systems for recognition and close
Companies building AI solutions need billing infrastructure that evolves with pricing strategy rather than constraining it. With 37% planning pricing changes within the next 12 months, the ability to test new models, apply retroactive changes, and maintain financial accuracy across complex scenarios is not a back-office concern. It determines how fast a company can respond to a market that is still repricing itself.
Frequently asked questions
What are the most common pricing models for AI agents?
Hybrid pricing dominates. Orb's 2026 study of 80 AI agent companies found 95% use hybrid pricing, 91.3% use usage-based pricing, and 71.3% use subscriptions, with 94.7% of the subscription users pairing subscriptions with usage charges. Outcome-based pricing appears in only 3.8% of the sample, down slightly from 4.5% in 2025, making it far less common than hybrid, usage-based, or subscription pricing in this sample.
How do AI agent vendors define a billable unit?
Deloitte describes pricing across actions and tasks, compute, API calls, tokens, duration, or combinations. In practice, published examples include resolutions at $0.99 from Intercom, resolved interactions at $0.90 from Gorgias, verified automated resolutions at Zendesk, whose cited tier documentation does not state a list rate, and Orb's credit abstraction model for tokens, agent runs, and outcomes.
What billing infrastructure do companies need for AI agent pricing?
AI agent billing requires real-time usage metering over a preserved layer of raw usage events, support for layered subscription plus usage structures, dimensional pricing across regions, instance types, and environments, credit systems for prepaid consumption, spend thresholds and alerts, and finance workflows that connect metered usage to ASC 606-aligned reporting and the ERP. Because pricing can change materially after launch, the system also needs versioning, simulation against historical usage, and backdating so corrections can be applied without manual reconstruction.
How is AI agent pricing different from traditional SaaS billing?
Traditional SaaS relies on predictable seat-based or flat subscription pricing. AI agents introduce variable consumption driven by tokens, API calls, task complexity, and compute, and billing systems may apply different rates by dimensions such as region, instance type, or environment. Orb's 2026 research also identifies effort-based pricing, where compute time, steps, or task complexity become the usage measure. Outcome models add a further layer by charging for results rather than resources. Implementations may require attribution rules and billing caps, and can include no-charge rules when an outcome is not achieved.
How often should AI agent pricing be expected to change?
There is no universal cadence. ICONIQ found that 37% of companies plan to change their AI pricing model within the next 12 months, and Orb's 2026 research describes AI agent pricing as a continuously iterated system rather than a one-time packaging decision. Cost pressure reinforces the need for reassessment: Gartner projects that generative AI cost per resolution for customer service will exceed offshore human agent costs by 2030. That makes recurring pricing review and tooling for controlled changes an ongoing operating requirement.