Decagon Pricing

Decagon uses usage-based pricing for its enterprise AI customer-service platform rather than publishing traditional software tiers. Its commercial model supports both per-conversation and per-resolution pricing, while exact dollar rates are provided through custom quotes.

Enterprise buyers can prepare for a Decagon evaluation by understanding what each billing unit measures, which variables can affect total cost, and how different pricing models behave across changing support volumes and channel mixes.

Key Takeaways

• Decagon does not publish a public dollar rate card. Enterprise pricing requires a custom quote
• Two usage models are available. Decagon offers both per-conversation and per-resolution pricing
• Per-conversation is the more common Decagon model. The company says most customers gravitate toward this structure
• Volume and channel mix can affect pricing. Buyers should model expected interaction levels and confirm how each channel is treated in their specific quote
• Maven AGI provides a published starting reference. AWS Marketplace currently lists a 12-month Maven AGI contract starting at $25,000, with final pricing based on expected scope, complexity, and resolution volume

How Decagon Pricing Works

Decagon offers two ways customers can pay for its AI agents: per conversation and per resolution.

Per-Conversation Pricing

Under the per-conversation model, the customer pays a negotiated rate for each incoming conversation handled by the platform.

Decagon says higher-volume commitments can receive different rates and that most of its customers currently choose this model.

The main budgeting advantage is straightforward volume modeling. If an organization knows its approximate annual interaction volume, it can estimate usage by applying the negotiated conversation rate.

A basic model looks like this:

Annual usage cost = eligible conversations × contracted rate

Any fixed commitments, implementation charges, channel-specific costs, or other contractual components should then be added separately if they appear in the quote.

Per-Resolution Pricing

Under per-resolution pricing, the customer pays a higher fixed rate for conversations that are fully resolved, while escalated conversations are not charged as resolutions.

Larger resolution commitments can reduce the negotiated unit rate.

This model makes the definition of a resolution particularly important. Buyers should establish in writing:

• What qualifies as fully resolved
• Whether customer confirmation affects completion
• How abandoned conversations are treated
• Whether repeat contacts affect the original resolution
• How partial workflow completion is counted
• How reopened issues are handled
• How human escalations are excluded from resolution billing

These details allow finance and CX teams to model the same unit the vendor will ultimately use for invoicing.

Does Decagon Publish Its Actual Prices?

Decagon does not currently publish standard dollar pricing tiers or a public rate card.

That means there is no official public Decagon list price for:

• Per-conversation usage
• Per-resolution usage
• Platform access
• Voice usage
• Implementation
• Integrations
• Minimum annual commitments
• Enterprise support

For budgeting purposes, enterprise buyers should request a written quote based on their channel mix, annual interaction volume, automation scope, and integration requirements.

What Can Affect a Decagon Quote?

Usage volume is only one input into enterprise AI pricing.

Decagon's platform spans chat, voice, email, SMS, integrations, workflow execution, testing, analytics, and AI-agent management. Commercial scope can therefore vary considerably between deployments.

Conversation Volume

Higher interaction volumes create larger usage commitments.

Decagon says higher-volume commitments can affect negotiated per-unit rates, so buyers should model both average and peak periods.

Useful scenarios include:

• Normal monthly volume
• Peak-season volume
• Expected annual total
• High-growth scenario
• Lower-than-forecast scenario

This provides a clearer view of how a usage-based contract may behave as demand changes.

Channel Mix

Different customer-service channels have different underlying operating characteristics.

Decagon says communication channel and volume can affect pricing, with voice interactions often costing more to handle than chat. Buyers should confirm how channel mix is reflected in their specific quote.

Conversation Versus Resolution Volume

The same support operation can produce materially different billable volumes depending on the pricing model.

Suppose an organization receives 100,000 eligible conversations and the AI fully resolves 75,000.

Under a per-conversation contract, the commercial unit is based on the 100,000 conversations.

Under a per-resolution contract, the relevant usage volume would instead be based on the 75,000 interactions that satisfy the contract's definition of resolution.

The example shows how billable volume changes between the two models; contracted rates remain specific to each agreement.

Integrations and Workflow Scope

Decagon connects AI agents with CRMs, help desks, contact-center systems, knowledge sources, and other enterprise applications.

It supports pre-built integrations, APIs, MCP connectivity, and custom endpoints for retrieving data, triggering actions, and supporting escalations.

Its Agent Operating Procedures, or AOPs, can support workflows such as:

• Processing refunds
• Updating account information
• Looking up customer data
• Troubleshooting issues
• Triggering workflows
• Applying business rules
• Escalating to human support

Buyers should determine whether implementation, integration, or custom workflow work is included in their commercial proposal or priced separately.

Teams deciding between internal development and a platform can also use Maven's build-versus-buy framework to account for implementation and operating requirements beyond the quoted contract price.

What Buyers Receive Beyond the Billing Unit

Price comparisons become more useful when the scope of the underlying platform is clear.

Decagon includes capabilities for building, operating, testing, observing, and improving AI customer-service agents.

Agent Operating Procedures

AOPs let teams describe workflows through natural-language instructions backed by structured logic.

They support customer-service reasoning and actions in connected systems while giving teams visibility into agent behavior and workflow execution.

Omnichannel Support

Decagon supports customer interactions across channels including:

• Chat
• Voice
• Email
• SMS

The platform is designed to apply business logic and workflows across supported customer-service surfaces.

Integrations

Decagon provides pre-built integrations alongside API, MCP, and custom connectivity for enterprise systems.

Integration categories include:

• CRM
• Help desk and ticketing
• Knowledge systems
• Contact-center infrastructure
• Telephony
• Business applications

The platform can use these connections for information retrieval and supported actions.

Testing and Observability

Decagon provides capabilities for testing, versioning, conversation analysis, monitoring, and agent improvement.

Its AOP framework also gives teams visibility into workflow logic and execution.

These capabilities belong in a total-cost evaluation alongside the commercial billing unit.

Security and Compliance Considerations

Security requirements can affect both procurement and implementation scope.

Decagon lists the following security and compliance credentials:

• SOC 2 Type II
• ISO/IEC 27001:2022
• PCI DSS 4.0.1
• HIPAA
• GDPR
• CCPA
• EU AI Act-related compliance
• Penetration-testing documentation

These represent different forms of certification, audit, regulatory alignment, and security assurance rather than interchangeable credentials.

Buyers should confirm which reports, certifications, assessments, contractual commitments, and controls apply to their specific deployment.

The NIST AI Risk Management Framework provides a vendor-neutral reference for evaluating governance and AI risk alongside vendor-specific security requirements.

How Maven AGI Pricing Is Structured

Maven AGI uses resolution volume as a central pricing input.

The current AWS Marketplace listing shows a 12-month contract starting at $25,000.

AWS describes that figure as a minimum rather than a fixed enterprise tier. Final pricing depends on:

• Expected scope
• Complexity
• Volume of resolutions

AWS defines a Maven resolution as a customer query the AI agent handles end to end without human intervention.

That can include workflows such as:

• Validating identity
• Checking an order
• Processing a refund
• Answering a coverage question

The listing also distinguishes completed AI resolutions from cases that require human escalation.

Maven also provides information about its AWS Marketplace availability for enterprises that use existing cloud procurement channels.

Pricing Across Channels

Maven's AI agent platform uses one reasoning engine across chat, email, voice, and web.

The AWS Marketplace listing describes pricing in terms of overall resolution volume. Buyers should still confirm the exact channels, workflows, and commercial scope included in their agreement.

Integrations and Actions

Maven documents 100+ out-of-the-box integrations, with additional connectivity available through APIs and Model Context Protocol.

These integrations allow the agent to combine customer context and enterprise knowledge with approved actions in connected systems.

Configuration

Agent Designer gives CX teams tools for simulation, testing, evaluations, behavior configuration, and monitoring.

This matters for total-cost analysis because operating an AI agent involves ongoing testing, knowledge improvement, and workflow management in addition to the commercial usage charge.

Security and Governance

Maven's trust and compliance framework documents:

• ISO/IEC 42001 certification
• ISO/IEC 27001 certification
• ISO/IEC 27701 certification
• ISO/IEC 27017 certification
• ISO/IEC 27018 certification
• SOC 2 Type II audit
• PCI DSS v4.0 Level 1 Service Provider validation
• Independent HIPAA/HITECH assessment
• Independent GDPR assessment
• Independent CCPA/CPRA assessment

These credentials provide different forms of security, privacy, payment, and AI-governance assurance.

Decagon and Maven AGI Pricing: What Changes in the Cost Model?

Decagon uses usage-based pricing, while Maven AGI's published AWS Marketplace offer uses contract pricing shaped by expected scope, complexity, and resolution volume.

Decagon supports both per-conversation and per-resolution pricing, while its actual dollar rates remain quote-based. Maven AGI has a published starting reference of $25,000 for a 12-month contract through AWS Marketplace, with final pricing customized to the deployment.

For buyers, the difference changes what needs to be forecast.

With per-conversation pricing, the primary usage variable is the number of eligible interactions.

With per-resolution pricing, buyers also need a clear contractual definition of a successful resolution and an estimate of how many eligible interactions the AI will complete autonomously.

Neither structure should be evaluated from the billing unit alone. Channel mix, integrations, implementation requirements, support, workflow complexity, and contractual commitments can all affect total cost.

How to Model Total Cost of Ownership

A useful AI-agent cost model should include more than the quoted usage rate.

Maven's enterprise AI evaluation framework provides additional criteria for assessing commercial structure alongside resolution definitions, integrations, governance, testing, and production performance.

1. Establish Current Support Volume

Start with:

• Annual conversations or tickets
• Monthly seasonality
• Channel distribution
• Growth rate
• Average number of contacts per customer issue

2. Separate Eligible From Ineligible Work

Not every support interaction will necessarily be suitable for autonomous resolution.

Classify the workload into:

• Suitable for autonomous handling
• Suitable for AI assistance with human review
• Requires human judgment
• Restricted by policy or regulation

Human agents remain central to sensitive conversations, complex exceptions, relationship-building, and work requiring judgment or empathy.

3. Model the Contract's Billing Unit

For per-conversation pricing:

Projected billable volume = eligible conversations

For per-resolution pricing:

Projected billable volume = eligible conversations × expected autonomous resolution rate

The resolution assumption should come from testing against the organization's actual issue mix rather than a vendor-wide benchmark.

4. Include Fixed and Variable Costs

Request clear treatment of:

• Platform commitments
• Usage fees
• Implementation
• Integrations
• Voice or telephony costs
• Support
• Professional services
• Overage pricing
• Contract minimums
• Renewal terms

The quoted unit rate represents only one part of total cost.

Architecture can create additional operating considerations as well. Maven's analysis of the cost of fragmented AI examines the operational overhead that can emerge when customer-service intelligence is distributed across disconnected systems.

5. Compare Against Business Outcomes

Cost per resolved issue can provide more context than raw cost per conversation.

Other useful measures include:

• Autonomous resolution
• First-contact resolution
• Time to resolution
• Repeat-contact rate
• Escalation rate
• Customer satisfaction
• Cost per ticket
• Human-agent time spent on escalated cases

Maven's guide to resolution versus deflection explains why keeping an interaction away from a human queue is different from completing the customer's underlying request.

Questions to Ask Before Signing an AI Agent Contract

Enterprise buyers should request enough detail to reproduce the vendor's invoice calculation internally.

Useful questions include:

• What exactly creates a billable conversation?
• What exactly qualifies as a billable resolution?
• Are human escalations charged?
• Is pricing different for voice, chat, email, or SMS?
• Are implementation services included?
• Are integrations included?
• Are custom integrations priced separately?
• Is there an annual minimum?
• How are overages calculated?
• Do unit rates change at higher volumes?
• How are abandoned interactions counted?
• How are reopened cases handled?
• What happens if support volume exceeds the committed forecast?
• What renewal adjustments should buyers expect?

Clear answers make pricing comparisons more consistent even when vendors use different commercial models.

Why Maven AGI Fits Enterprise Pricing Evaluations

Maven combines a published AWS Marketplace starting point with pricing tied to the expected scope, complexity, and volume of completed resolutions. This gives enterprise buyers a concrete starting reference while keeping final commercial terms aligned with the workflows being deployed.

The platform also connects pricing considerations with the broader operating environment. Maven's 100+ integrations connect with existing help desks, CRMs, knowledge systems, telephony platforms, and enterprise applications, while Agent Designer gives CX teams tools to test, evaluate, monitor, and improve agent behavior over time.

Maven's shared reasoning layer extends across supported customer channels, allowing the same enterprise knowledge, policies, and approved actions to support chat, email, voice, and web experiences. Human escalation remains part of the operating model when judgment, empathy, or complex exception handling is required.

Documented customer outcomes also provide useful context for evaluating resolution economics. Papaya reports 90% of chat inquiries answered autonomously, 70% first-contact resolution, and a 50% reduction in cost per ticket after deploying Maven. These results reflect Papaya's individual environment and implementation rather than a universal expected outcome.

For enterprise teams comparing AI-agent economics, Maven brings pricing, autonomous resolution, integrations, CX-team configuration, and enterprise governance into one evaluation framework.

Frequently Asked Questions

Does Decagon publish pricing?

Decagon does not currently publish standard dollar rates or pricing tiers. Its commercial pricing is quote-based, with per-conversation and per-resolution structures available. Buyers need a custom quote to determine the rates and commercial terms for their expected volume, channels, and deployment scope.

How do Decagon's per-conversation and per-resolution models differ?

Per-conversation pricing charges for incoming conversations handled by the AI, while per-resolution pricing charges for interactions that meet the contracted definition of a completed resolution. Decagon says most customers currently choose per-conversation pricing, although both options are available.

What should enterprise buyers confirm in a Decagon quote?

Buyers should confirm the billing unit, volume commitments, channel-specific treatment, implementation scope, integration costs, overage terms, and how escalations, abandoned conversations, and reopened cases affect billing. For per-resolution agreements, the contract should clearly define what qualifies as a completed resolution.

How does Maven AGI connect pricing with customer resolution?

Maven uses resolution volume as a central pricing input. Its AWS Marketplace offer defines a resolution as a customer query handled end to end without human intervention, while final pricing reflects expected scope, complexity, and resolution volume. This gives enterprise teams a commercial framework tied to completed customer-support outcomes.

What does Maven AGI offer beyond its pricing model?

Maven combines its pricing approach with a broader AI agent platform that connects enterprise knowledge, approved actions, and customer context across supported channels. Teams also have access to 100+ integrations and Agent Designer for testing, evaluating, monitoring, and improving agent behavior over time.