Chatbot Pricing Comparison: The Four Billing Models

Chatbot pricing ranges from zero dollars to several thousand dollars per month depending primarily on which of four incompatible billing units a vendor selects: message credits, billable conversations, per-seat software licenses, or automated outcome fees.

At Layer3Labs, we build and run custom automated workflows for client intake and customer support, and we repeatedly see teams misjudge their monthly software spend by evaluating entry-tier sticker prices instead of volume-scaling units.

Software vendor pricing changes without notice, so verify current rates on each vendor's pricing page before making purchasing decisions. This guide compares the published rates of Chatbase, Tidio, Intercom, Botpress, and Zendesk to show how each pricing structure behaves as customer support volume grows.

Self-Serve Platform Pricing vs. Support Suite and Per-Outcome Pricing: Side-by-Side

Suggest a correction — if you work at one of the products above and something here is out of date, tell us and we'll fix it.

The Four Incompatible Chatbot Pricing Units

Comparing chatbot pricing requires translating software quotes into one of four incompatible billing units: message credits, billable conversations, per-seat licenses, or outcome-based resolution fees. Evaluating these platforms purely on their entry-level monthly subscription cost leads to severe miscalculations because a forty-dollar plan on one billing unit can easily generate higher operating invoices than a hundred-dollar plan on another.

The first model is message-credit billing, used by tools like Chatbase. In this system, every single back-and-forth interaction consumes an allowance. If a customer asks a question, the bot clarifies, the customer answers, and the bot provides a resolution, that single inquiry has consumed multiple message credits. When credits run out, the bot either stops responding or charges overage rates.

The second model is conversation-based billing, used by platforms like Tidio and Botpress. Here, a conversation represents an entire session or inquiry thread within a defined time window, regardless of whether the user exchanges two messages or twelve messages with the bot. This model offers greater predictability for complex support questions that require back-and-forth diagnosis.

The third model combines per-seat software access with per-outcome AI fees, exemplified by Intercom. You pay a baseline fee for every human support agent using the helpdesk inbox, plus a variable charge each time the automated bot completes a defined task or resolves a conversation. The fourth model, seen in Zendesk, couples per-agent software tiers with automated resolution fees that trigger only when an issue is closed without human agent intervention.

• Message credits: Best for simple, single-turn FAQ lookups; expensive for multi-turn diagnostic support.
• Billable conversations: Predictable unit economics when customer inquiries require extensive back-and-forth clarification.
• Per-seat plus per-outcome: Aligns software expenditure with successful ticket resolutions while maintaining a baseline seat license cost.
• Per-seat plus automated resolutions: Enterprise helpdesk model where automated containment offsets human staffing requirements.

→ Compare per-token pricing for every AI model

1. Chatbase: Message Credits and Seat Add-Ons

Chatbase structures its pricing around monthly message credit allowances and workspace seats rather than unique customer conversations. The platform ingests website links, documents, and text files to generate an automated question-answering bot trained on proprietary company data.

On the Free tier, Chatbase provides 50 message credits per month, 1 agent, and 1 workspace seat at zero cost. The Hobby tier costs $40 per month and includes 700 message credits along with 2 workspace seats. The Standard tier, marked as popular by the vendor, costs $150 per month for 4,000 message credits and 3 workspace seats. The Pro tier reaches $500 per month, granting 15,000 message credits and 5 workspace seats. Enterprise plan pricing is not shown on the page.

Chatbase advertises a 20% discount on yearly plans, but specific annual prices are not displayed on its pricing table. Specific agent counts for paid tiers are also not shown on the page. We publish a full Chatbase pricing guide that analyzes these credit limits in detail. Vendor pricing changes without notice, so verify current rates on the Chatbase pricing page before purchasing.

The practical consequence of this pricing unit is that multi-turn conversations drain credit balances rapidly. If an average customer support inquiry involves three customer questions and three bot answers, that single customer consumes six message credits. Under that usage pattern, the $40 Hobby plan supports approximately 116 customer interactions per month before exhausting its credit quota.

• Free tier: $0 per month, 50 message credits per month, 1 agent, 1 workspace seat.
• Hobby tier: $40 per month, 700 message credits per month, 2 workspace seats.
• Standard tier: $150 per month, 4,000 message credits per month, 3 workspace seats.
• Pro tier: $500 per month, 15,000 message credits per month, 5 workspace seats.
• Enterprise tier: Custom pricing not shown on the public pricing page.

2. Tidio: Billable Conversations and Lyro Add-Ons

Tidio bills customer service teams on total monthly billable conversations while treating autonomous generative AI responses as an add-on service. The platform integrates a shared live-chat widget, rule-based automation flows, and its generative AI agent called Lyro into a unified inbox.

The Free plan on Tidio provides 50 billable conversations per month at zero cost. The Starter plan begins at $24.17 per month for 100 billable conversations, 100 Flows visitors reached, and a one-off allocation of 50 Lyro AI conversations. The Growth tier starts from $49.17 per month for 250 billable conversations. The Plus tier starts from $300 per month plus monthly usage, offering custom conversation limits and discounted usage rates. The Premium tier requires contacting sales, provides custom limits, starts from 3,000 Lyro AI conversations, includes a guaranteed 50% Lyro AI resolution rate, and offers a pay-per-resolution billing option.

For teams scaling their automated support, Tidio sells standalone add-on packages. The Lyro AI Agent add-on starts from $32.50 per month for 50 conversations, while rule-based Flows start from $24.17 per month for 2,000 visitors reached. In the open-source ecosystem, Tidio ships a plugin listed in the WordPress plugin directory with 70,000+ active installations and 396 ratings, listed as free to add. Vendor pricing changes without notice, so verify current rates on the Tidio pricing page before buying.

Because Tidio separates standard live-chat conversations from Lyro AI interactions, a business handling 500 total monthly chats must budget for both the base subscription tier and supplementary Lyro conversation packs once the initial promotional allowance expires.

• Free tier: $0 per month, 50 billable conversations per month.
• Starter tier: $24.17 per month, 100 billable conversations, 50 one-off Lyro conversations, 100 Flows visitors.
• Growth tier: From $49.17 per month, starting at 250 billable conversations.
• Plus tier: From $300 per month plus usage, custom conversation limits, discounted overages.
• Lyro AI Add-on: Starts from $32.50 per month for 50 autonomous conversations.

3. Intercom: Seat Subscriptions and Fin Per-Outcome Fees

Intercom combines human seat subscriptions for support representatives with a flat ninety-nine cent charge for every autonomous resolution completed by Fin AI. This hybrid model separates the software infrastructure required by human agents from the variable automation delivered by machine learning models.

Base helpdesk subscriptions on Intercom require per-seat commitments: the Essential plan costs $29 per seat per month, the Advanced plan costs $85 per seat per month, and the Expert plan costs $132 per seat per month. Across all tiers, the Fin AI Agent costs exactly $0.99 per Fin outcome. Intercom also sells Fin as a standalone product to integrate with existing external helpdesks, though its standalone per-seat software price is not shown on the page. Vendor pricing changes without notice, so verify current figures on the Intercom pricing page before contracting.

Evaluating Intercom requires understanding its exact definition of an outcome. In the Intercom blog post From resolutions to outcomes, Intercom defines an outcome as when Fin successfully completes the action it was configured to perform as part of a conversation. This includes full issue resolutions and Procedure cases where Fin collects structured customer data and routes the ticket to a human. An outcome is therefore a broader billable unit than an autonomous ticket resolution. Intercom reports its average customer resolution rate stands at 76%.

Conversely, marketing materials on the Fin benchmarks page, covering 110M+ conversations across 12,000+ customers and 15 industries, publish top-10 averages rather than overall customer medians: an 85% resolution rate, a 91% involvement rate, and a 78% automation rate. CSAT and response times are not published on that benchmark page. Buyers must note that 85% represents top-decile performance rather than a typical baseline.

• Essential plan: $29 per human seat per month.
• Advanced plan: $85 per human seat per month.
• Expert plan: $132 per human seat per month.
• Fin AI Agent usage fee: $0.99 per Fin outcome across all subscription tiers.
• Outcome definition: Includes both resolved issues and structured information collection handoffs.

4. Botpress: Developer Workflows and Bundled Execution Spend

Botpress offers an open developer architecture that prices bot deployments based on monthly active conversations with bundled model execution spend. Unlike consumer chat widgets, Botpress provides visual flow programming, API execution steps, and integration connectors suited for software engineers.

The Free plan on Botpress costs $0 per month and includes 25 conversations per month. The Plus plan costs $150 per month billed annually, which provides 250 conversations per month and allows buying extra packs of 100 conversations for $65 each. The Team plan costs $750 per month billed annually, including 1,500 conversations per month with additional packs of 100 conversations priced at $50 each. Enterprise plans carry custom pricing for custom conversation volumes.

Botpress bundles artificial intelligence execution costs directly into its plans. The vendor specifies that every conversation includes $0.10 of AI usage covered by Botpress. That equates to $25 of included AI spend on the Plus tier and $150 of included AI spend on the Team tier. Vendor pricing changes without notice, so verify current terms on the Botpress pricing page before planning your architecture.

For teams with internal developers, Botpress eliminates per-seat fees entirely. A customer support team of twenty representatives can access Botpress without paying seat software premiums, paying only for the underlying conversation volume and automated workflow executions.

• Free tier: $0 per month, 25 conversations per month.
• Plus tier: $150 per month billed annually, 250 conversations included, $65 per extra 100 conversations.
• Team tier: $750 per month billed annually, 1,500 conversations included, $50 per extra 100 conversations.
• Enterprise tier: Custom pricing tailored for high-volume deployments.
• AI spend inclusion: Includes $0.10 of AI model spend per conversation covered by the vendor.

5. Zendesk: Helpdesk Seats and Automated Resolutions

Zendesk pairs traditional enterprise helpdesk seat licenses with usage-based billing tied to autonomous ticket resolutions. The platform is designed for large support departments that manage customer communications across email, phone, web forms, and real-time chat widgets.

Base helpdesk access on Zendesk starts with Support Team at $19 per agent per month paid yearly. The unified Suite Team plan costs $55 per agent per month paid yearly, while Suite Professional costs $115 per agent per month paid yearly. The Suite Enterprise plus Copilot plan requires talking to sales, and its price is not shown on the page. Zendesk also sells a Copilot add-on for $50 per agent per month paid yearly. Monthly, non-annual subscription rates are not shown on the pricing table.

Artificial intelligence agents are included across every Suite and Support plan, with pricing based on the successful outcomes they deliver. Zendesk calls its billing unit an Automated Resolution. However, the exact per-resolution price is not shown on the public pricing page. Vendor pricing changes without notice, so verify current figures on the Zendesk pricing page before signing a contract.

Budgeting for Zendesk requires calculating two separate variables: the number of support agents requiring monthly platform seats, and the expected volume of customer tickets that the automated AI bot will successfully close without human intervention.

• Support Team tier: $19 per agent per month paid yearly.
• Suite Team tier: $55 per agent per month paid yearly.
• Suite Professional tier: $115 per agent per month paid yearly.
• Copilot Add-on: $50 per agent per month paid yearly.
• Automated Resolution billing: Included in plans, but the specific fee per resolution is unpublished.

Unpublished Line Items and Opaque Tier Pricing

Evaluating enterprise chatbot proposals requires identifying multiple critical fees that vendors deliberately omit from public pricing tables. These hidden or unquoted line items frequently determine whether a software deployment remains within budget as operational scale increases.

Across the platforms examined, several critical pricing components remain entirely unpublished. Chatbase does not publish Enterprise tier pricing, specific annual plan dollar amounts, or agent limits for paid tiers. Intercom does not publish the per-seat price for using Fin standalone on external helpdesks. Zendesk does not publish its price per Automated Resolution, its monthly non-annual billing rates, or the price of its Suite Enterprise plus Copilot tier. Tidio requires direct sales contact for its Premium tier and negotiates custom limits on both Plus and Premium.

Third-party messaging channels also introduce separate, mandatory cost structures. Our HTTP request to the ManyChat pricing page returned an HTTP 403 response on 2026-09-22, meaning verified live rates cannot be confirmed here; consult our dedicated ManyChat pricing guide and the official ManyChat pricing page directly. ManyChat specializes in direct message automation across Instagram, Facebook Messenger, and WhatsApp rather than website search widgets.

For teams deploying automated chatbots on WhatsApp, the underlying infrastructure costs changed fundamentally on 1 July 2025. As documented on the official WhatsApp Business Platform pricing portal, WhatsApp transitioned from conversation-based billing to a per-message model. Fees apply whenever a Marketing, Utility, or Authentication template message is delivered outside the standard customer service window. Per-message rates are country-specific and are not listed here; consult the WhatsApp rate cards directly.

• Chatbase: Enterprise pricing and specific annual dollar figures are unpublished.
• Intercom: Standalone Fin helpdesk seat licensing rates are not publicly listed.
• Zendesk: Cost per Automated Resolution and non-annual monthly rates are omitted from the site.
• Tidio: Premium tier volume pricing and Plus overage rates require custom sales quotes.
• WhatsApp API: Deprecated conversation pricing in favor of per-message template fees starting July 2025.

→ Compare per-token pricing for every AI model

Custom Build Engagements and Agency Cost Structures

Contracting an agency or systems integrator to build a bespoke AI chatbot shifts capital expenditure from recurring software seat markups into upfront architecture and integration engineering. While self-serve software vendors charge monthly subscriptions for prebuilt widgets, bespoke development creates dedicated workflows connected directly to internal company databases and APIs.

Agency pricing follows a predictable four-stage cost shape: discovery and architecture planning, integration and data pipeline engineering, user interface deployment, and ongoing system maintenance. During discovery, engineers audit knowledge repositories, define escalation protocols, and benchmark prompt performance. The integration phase connects the bot to company CRMs, inventory systems, or practice management tools.

At Layer3Labs, we find that organizations pursue custom engineering when data governance requirements prohibit sending customer records to multi-tenant software platforms, or when automated interactions require executing transactional database read-and-write operations rather than simple document search. Read our operational guide to AI Chatbot Development for Small Business to evaluate whether custom engineering suits your technical requirements.

Ongoing costs for custom builds include vector database hosting, language model API token consumption, and monthly maintenance retaining an engineer to monitor response quality. Compare these operational overheads against full-time staffing expenses using our detailed model on Chatbot vs Human Agent Cost.

• Discovery and prompt engineering: Scope architecture, verify security posture, and clean documentation.
• System integration: Build API webhooks into CRMs, billing systems, and customer ticketing databases.
• Testing and guardrails: Run evaluation datasets to confirm factual accuracy and eliminate hallucinations.
• Ongoing maintenance: Manage API token usage, refresh knowledge sources, and update integration endpoints.

→ Compare per-token pricing for every AI model

Quote Normalization via Cost Per Resolved Conversation

Comparing competing chatbot quotes accurately requires converting all seat fees, message quotas, and usage overages into an estimated cost per resolved customer interaction. Because vendors bill on message counts, conversation sessions, or closed tickets, direct plan-to-plan price comparisons are mathematically meaningless without a common denominator.

To normalize quotes across vendors, apply a standard evaluation formula: calculate your total monthly expenditure across seat licenses, base plans, and expected overage fees, then divide that sum by the total number of inquiries successfully resolved by the software. You can test your organization's specific support volume using our interactive support savings calculator.

Consider a business with two human support agents handling 500 incoming customer inquiries per month, where an automated bot resolves 50% of the volume (250 inquiries), leaving 250 inquiries for human agents. On Chatbase, assuming each conversation takes five back-and-forth messages, those 500 inquiries consume 2,500 message credits. The business must purchase the $150 Standard plan (4,000 credits, 3 seats), yielding a cost of $0.60 per resolved conversation ($150 divided by 250 resolutions).

On Intercom, two Essential seats cost $58 per month ($29 multiplied by two). The 250 resolved customer inquiries cost $247.50 ($0.99 multiplied by 250 outcomes), assuming every resolution meets Intercom's outcome criteria. The total monthly cost is $305.50, yielding a normalized cost of $1.22 per resolved conversation. While Intercom's entry seat price appears cheaper than Chatbase's Standard plan, its per-outcome pricing makes it more expensive at this specific volume.

• Step 1: Estimate total monthly customer inquiry volume and average messages per inquiry.
• Step 2: Determine your expected autonomous resolution rate based on documentation quality.
• Step 3: Add human agent seat licenses to base software subscription costs.
• Step 4: Calculate consumption fees for message credits, extra conversation packs, or outcome charges.
• Step 5: Divide total monthly projected expenditure by total autonomous resolutions.

→ Compare per-token pricing for every AI model

Deployment Conditions Favoring Alternative Support Channels

Deploying an autonomous chatbot produces negative returns when customer ticket volume is minimal or when inquiries demand discretionary human judgment. Organizations handling fewer than 50 customer inquiries per month rarely justify paid chatbot software, because human staff can resolve incoming tickets using free shared inboxes without incurring monthly software overhead.

High-liability environments also represent a poor fit for autonomous chatbot resolution. Legal practices handling active litigation, medical clinics triaging acute symptoms, and crisis hotlines cannot afford the hallucination risks inherent in generative language models. In these sectors, automated systems should restrict their scope to structured intake forms that collect contact information and route tickets directly to human staff.

Our recommendation would flip if an organization with unorganized internal documentation experiences a sudden surge in repetitive support tickets. If that business first consolidates its policies into an accurate, indexed knowledge base, an outcome-based bot like Intercom Fin or an automated builder like Chatbase can start absorbing the repetitive share of that volume. Until internal documentation is structured and verified, investing in automated chat software will amplify errors rather than reduce support costs.

Audit your support inbox for thirty days to catalog your monthly conversation volume, average back-and-forth messages per inquiry, and total required human seats before signing any vendor contract.

• Low-volume operations: Under 50 monthly inquiries, free tools or manual human email handling are more cost-effective.
• High-liability domains: Inquiries requiring regulatory compliance or legal accountability require direct human review.
• Disorganized documentation: Automated bots reproduce factual errors when company documentation contains conflicting information.
• Transaction-heavy support: Platforms requiring complex database modifications require custom API engineering rather than basic chat widgets.

The Verdict

Do not select a chatbot platform based on headline monthly subscription fees. A $40 platform that charges for message credits can rapidly outprice a $150 plan once customer interactions expand past basic two-turn questions.

Choose Chatbase if you need an accessible, document-trained site widget and your team has zero human support agents requiring helpdesk software seats. It provides the most straightforward pricing for public website FAQ deflection, provided you monitor credit consumption carefully.

Choose Tidio if you run a small e-commerce business on WordPress or Shopify and require a combined solution for live chat, rule-based marketing flows, and generative AI support. Its free tier and affordable Starter plan allow small shops to experiment before committing to higher-volume Lyro AI conversation packs.

Choose Botpress if you have software developers who can build custom conversational flows and integrate backend APIs. By eliminating per-seat software fees and bundling $0.10 of AI spend per conversation, it delivers exceptional cost efficiency for technical teams managing high inquiry volumes.

Choose Intercom or Zendesk if your business operates an established customer support department with multiple human agents, complex ticketing workflows, and high resolution standards. You will pay higher base seat fees and variable per-outcome charges, but the tight integration between automated bots and human escalation prevents dropped customer interactions.

If your inquiry volume is under 50 tickets per month, stay on manual email or basic free live chat until volume justifies the operational setup. When you are ready to evaluate platforms, calculate your cost per resolved conversation using your own volume numbers rather than vendor marketing benchmarks.