Decagon Pricing: What It Really Costs in 2026

Decagon pricing is not published anywhere on decagon.ai. Decagon itself has confirmed, in its own product blog, that it quotes deployments one of two ways: per conversation (a fixed usage-based rate for every incoming chat, voice, or email exchange) or per resolution (a higher, outcome-based rate charged only when the AI fully resolves an issue without human help). Every account starts with a demo and a custom quote, not a price list.

That is a more specific answer than most "quote-only" vendors give, because Decagon has written publicly about how it thinks about its own pricing. Here is what is actually confirmed, what is third-party estimate, and how it compares to the rest of the enterprise AI customer-support category.

Decagon pricing at a glance

How does Decagon pricing actually work?

Decagon is refreshingly direct about this compared to most enterprise AI vendors: in a December 2024 post titled "Pricing the AI Agent Economy," Decagon's Director of Product wrote that the company supports two pricing structures for its customer-service agents. Per-conversation pricing charges a fixed rate for every incoming conversation, with volume discounts at scale. Per-resolution pricing charges a higher fixed rate, but only for conversations the AI fully resolves without escalating to a human, with no charge for the ones it hands off.

The detail that matters for budgeting: Decagon says the "vast majority" of its customers choose per-conversation pricing, not per-resolution. That runs against the common assumption that Decagon is purely an outcome-based, pay-per-resolution vendor. It offers that model, but usage-based billing is what most of its book actually runs on, according to Decagon's own account.

Is Decagon priced per conversation or per resolution?

Both are on the table, and which one you land on is a negotiation, not a fixed policy. Decagon's own reasoning for why customers lean toward per-conversation pricing is worth taking at face value, since it is unusually candid for a vendor's own blog: outcome-based billing creates constant arguments over what counts as a "resolution." If a frustrated customer abandons the chat, does that count? If the AI gives a technically correct but unhelpful answer, does that count? Decagon's own glossary entry on resolution-based pricing lists this ambiguity as the model's core weakness, alongside less predictable month-to-month billing.

Per-conversation pricing sidesteps that argument entirely: you pay for volume, full stop, and the incentive fights disappear. That is a genuinely useful piece of buyer education, and it is the kind of thing you will not find on a typical "contact us for pricing" page. For a broader look at how the category prices these two models, see our breakdown of Zendesk AI pricing, which layers per-resolution AI billing on top of seat pricing, and Intercom Fin pricing, which runs a similar per-resolution structure for its support agent.

Why doesn't Decagon publish a price list?

The short answer: none of the enterprise AI customer-support vendors do, and Decagon is not an outlier there. Sierra, Ada, and most of the platforms covered in our roundup of AI customer-service companies all route pricing through a sales conversation rather than a self-serve checkout. The reason is structural, not secretive: cost depends on ticket volume, which channels you deploy (chat, voice, email, or all three), how deep the integrations run into your existing helpdesk or CRM, and what resolution rate the AI actually hits in your environment. None of that is knowable before a vendor sees your data. If you are comparing the category, Sierra alternatives and Ada competitors are both worth a look, since the buying dynamic (custom quote, sales-led, enterprise-only) is nearly identical across all three.

What actually moves a Decagon quote up or down?

Based on Decagon's own pricing structure plus consistent third-party reporting, four levers explain most of the spread between a $105K contract and a $923K one.

Conversation volume

Both pricing models scale directly with how much the AI handles, so monthly ticket count is the single biggest input. Whether you land on per-conversation or per-resolution billing, volume is the multiplier every other factor sits on top of.

Per-conversation vs. per-resolution model

The billing model you pick changes the math, not just the label. Per-conversation pricing charges a flat rate for every incoming chat, voice, or email exchange, with volume discounts at scale. Per-resolution pricing charges a higher rate, but only for conversations the AI fully resolves without a human, so a higher AI resolution rate means a higher bill, which is exactly the alignment (and the risk) Decagon describes in its own glossary post.

Deployment complexity

Channel mix and integration depth move together here. Decagon sells chat, voice, and email as a unified platform, and adding voice in particular is reported to carry a premium over chat-only deployments. Separately, connecting Decagon to a standard helpdesk is one thing; wiring it into a legacy or custom system typically adds professional-services cost, per third-party buyer guides like Ringg AI's analysis.

Contract length

Multi-year commitments and guaranteed uptime or dedicated support typically pull the effective rate down or up depending on which way you negotiate. Longer terms buy leverage on the per-conversation or per-resolution rate; shorter ones buy flexibility.

None of these are published as a rate card. They are reported patterns from analysts and competitors who have watched enterprise AI CX deals close, not confirmed numbers from Decagon.

How does Decagon's contract size compare to other enterprise AI vendors?

Treat any specific dollar figure here as a third-party estimate, not a Decagon-confirmed price. With that caveat: one buyer-side analysis, My AskAI, estimates typical Decagon contracts run from roughly $105,000 to $923,000 per year, with a median near $433,000. That places Decagon in a similar band to other enterprise AI agent platforms with opaque pricing: our own research on Kore.ai pricing found reported enterprise deals starting around $300,000 a year, and Aisera pricing reporting a median closer to $108,000. Directionally, all of these platforms share the same shape: no public rate card, a sales-led process, and six-figure annual contracts once you are past a pilot. Our quarterly enterprise AI price index tracks this pattern across the category if you want the fuller picture.

An underused signal for reading where Decagon sits in that band is its own funding trajectory. None of the other pricing writeups on Decagon connect the two, but valuation and go-to-market focus tend to move together:

None of this is a Decagon-confirmed pricing signal, but it is a reasonable read of public funding data: a company tripling its valuation to fund enterprise sales capacity is not a company about to publish a self-serve price list. That trajectory also builds on an earlier signal: The Information reported in November 2025 that Decagon was already in talks to raise at a $4 billion valuation, months after the $1.5B Series C and just before the round closed closer to $4.5 billion. That is a company selling into large enterprise accounts, not a self-serve SaaS tool, and the contract sizes track accordingly.

Is resolution-based pricing actually good for buyers?

It depends on how disciplined your team is about defining success before you sign. Decagon's own glossary is candid that resolution-based pricing "rewards outcome delivery, not just usage," which sounds appealing until you hit the model's honest downside: someone has to define what a "resolution" is, in writing, before the meter starts running. Decagon lists the exact failure modes: a customer who abandons a chat mid-conversation, or an AI answer that is technically given but does not actually solve the problem. Get that definition wrong in the contract and you are relitigating your bill every month.

Per-conversation pricing avoids that fight but means you pay the same rate whether the AI nails the issue or completely misses it. Neither model is objectively better; they are a trade between predictability and outcome alignment, and Decagon's own team has said most of its customers value the predictability of per-conversation billing more than the outcome purity of per-resolution billing.

Should you use Decagon, or does your team need something different?

Decagon is a strong, well-funded choice if your problem is specifically external customer support: deflecting chat, voice, and email volume from human agents at scale. The $4.5 billion valuation and the customer-reported deflection numbers on its own site reflect real demand for that category, and it deserves credit for being unusually transparent about how its own pricing philosophy works, even without a public price list.

But Decagon solves one specific problem: conversations with your customers. It does not touch the work your own team does internally, like pulling context from Salesforce before a renewal call, drafting a Jira ticket from a Slack thread, or prepping a QBR from three months of scattered meeting notes. That is a different job, and it is the job Coworker AI is built for.

What does Coworker AI cost compared to Decagon?

Coworker AI is an internal AI teammate, not a customer-facing support agent, so the comparison is really "quote-only enterprise sales" versus "self-serve, published pricing." Coworker runs Free at $0, Pro at $29.99 per user per month, and Max at $149.99 per user per month, with an Enterprise tier for custom deployments. There is no sales call required to see the number, and no per-resolution meter to negotiate. Coworker connects to 50+ tools (Slack, Salesforce, HubSpot, Google Drive, Jira, GitHub, and more), and its OM1 organizational memory keeps context from all of them so it can actually execute tasks, not just answer questions about them, the same execution-over-search gap covered in our best AI agents roundup. If Decagon's six-figure sales cycle and per-conversation math are more than your internal workflows need, get started free and see what a flat, published rate looks like instead. You can also compare the full lineup on the pricing page or see what else is possible on the Coworker AI homepage.

Frequently asked questions

How much does Decagon cost? Decagon publishes no prices. It quotes deployments per conversation or per resolution based on volume, channels, and contract length, and third-party estimates put typical annual contracts between roughly $105,000 and $923,000, with a median near $433,000. None of those dollar figures are confirmed by Decagon itself.

Is Decagon's pricing outcome-based or usage-based? Both models exist. Decagon supports per-resolution (outcome-based) and per-conversation (usage-based) pricing, and Decagon's own blog states that most customers choose the usage-based per-conversation option over the outcome-based per-resolution option.

Does Decagon have a free trial? No self-serve trial or sign-up flow is visible on decagon.ai. The only conversion path on the site is "Get a demo," which leads to a sales conversation.

Is Decagon cheaper than Sierra? There is no way to say definitively, since neither Decagon nor Sierra publishes prices. Both are enterprise, quote-only vendors that price around outcomes and usage rather than seats. See our Sierra alternatives breakdown for how the two compare on positioning.

Does Decagon charge extra for voice support? It is not published, but third-party buyer guides like Ringg AI report that adding voice on top of chat and email typically carries a premium, consistent with how most omnichannel AI CX platforms price.

Is there a Coworker AI alternative to Decagon? Not a direct one. Decagon automates external customer support conversations; Coworker AI is an internal AI teammate that executes work across your connected tools with organizational memory. Teams sometimes run both: Decagon for customer-facing deflection, Coworker for the internal work behind the scenes.

Related reading

• The Enterprise AI Price Index
• AI customer service companies compared
• Sierra alternatives
• Ada competitors
• Zendesk AI pricing
• Intercom Fin pricing
• Kore.ai pricing
• Aisera pricing
• Best AI agents

For a similar outcome-based CX agent, compare Sierra AI pricing.