Chat Software Examples: Five Products, Three Different Meters

Why the meter matters more than the tier

Every chat product presents a tiered pricing table. The tiers look similar across vendors: a starter level, a team level, a business level with more features. You compare the column headers and pick the one that fits your team size.

That comparison is mostly meaningless, because the columns are priced on different meters. One vendor charges for every person who can log in. Another charges for every conversation started. Another charges for every question the software resolves without a human. The same monthly bill on a per-seat table and a per-conversation table can represent wildly different workloads, and only one of them is the number you actually care about.

The grouping below sorts them by meter. Read it with your own monthly volume in mind.

Starting prices across the five chat software examples, grouped by the meter each one uses.

Pricing model Vendor / Starting price Bar

Per agent

LiveChat from

$19

Olark live chat from

$29

Per conversation

Tidio from

$24.17

Per workspace

Crisp from

$0

Per seat

Intercom from

$29

Asyntai from

$39

Per resolved outcome

Intercom Fin AI from

$0.99

Bar lengths scaled to the highest starting price ($39). Crisp free tier has no bar. Olark AI Website Agent is $400 per month, separate from the live chat tier.

Five chat software examples sorted by pricing model

These are the chat software examples most frequently compared on pricing review sites. They fall into three groups by what drives the bill.

Chat software examples priced per agent or per seat

The most common pricing model. You pay for every login, regardless of how busy that person is. The model suits a staffed inbox with defined hours, because you know how many people are available and you pay for that headcount whether the queue is empty or not.

LiveChat

LiveChat charges per person, and the price is identical whether that person answers one chat or forty in a day. The annual billing discount is significant: $19 per person per month billed annually versus $25 billed monthly, which is a 24% saving on the Starter tier.

The Asyntai row is ours. Read it as a disclosure: we make Asyntai, so the figure is published here because it belongs to us, not because we are recommending it against LiveChat. LiveChat data from its own pricing page, checked September 2026.

Routing is spread across the ladder rather than held at the top. LiveChat's pricing page puts smart chat distribution between available agents on Starter, adds routing rules by URL and by geolocation on Team, and adds the work scheduler for shift handover on Business. So the step from Starter to Team is the one that buys conditional routing, not the step to Business.

Olark

Olark splits into two distinct products. The live chat product is straightforward per-agent pricing at $29 per agent per month. The AI Website Agent is a separate flat-rate product at $400 per month and includes the first seat. These are not add-ons to the same plan; they are two products with different billing structures, and a buyer who wants both pays both.

Olark data from its own pricing page, checked September 2026. The AI Website Agent price is a flat $400 per month billed annually ($4,800 per year); it includes the first seat. Compare this to LiveChat where the AI layer is a separate feature on the Business tier rather than a separate product.

Intercom

Intercom prices per seat, but adds a second meter on top: Fin AI, their AI layer, charges $0.99 per resolved outcome. The resolved outcome is a specific event: a conversation that the AI handles completely, without routing to a human. A conversation that starts with the AI and then hands to a human agent is billed differently, because the resolution happened differently.

Intercom data from its own pricing page, checked September 2026. Fin AI is priced per resolved outcome at $0.99 each, charged once per conversation that the AI resolves entirely. The Fin AI column shows the base platform cost; the add-on cost scales with your resolution volume and is the input you vary to calculate your own bill.

The per-seat tiers cover the core live chat and inbox functionality, and the feature split between them moves. Before you pick a tier, open Intercom's own plan comparison and check two things against your workflow: which tier carries the rules engine you need for routing, and which tier carries the reporting your team lead will actually open.

Chat software examples priced per workspace

Instead of charging per person, these charge per workspace. The workspace is the shared environment that contains conversations, contacts and configuration. You pay for the workspace and get a set number of seats included, with additional seats as an add-on.

Crisp

Crisp data from its own pricing page, checked September 2026. The Free tier includes 2 seats and 100 contact profiles. The Plus tier includes 20 seats in the base price; additional seats are $10 each. The lower tiers do not offer additional seat purchases; if you need more than the included seats, you move to the next tier.

Crisp is notable for a genuine free tier with two seats and a contact limit of 100 profiles. This is usable for a small business that wants to evaluate the product, though the profile limit catches up quickly on any site with meaningful traffic. The workspace model means that if a team exceeds its included seats, the next tier adds seats in a block rather than paying per individual.

For a 12-person team, the Plus tier at $295 includes 20 seats. A 12-person team at Crisp costs $295 per month. Crisp does not publish a per-message or per-conversation meter; the workspace is the unit.

Chat software examples priced per conversation

The conversation meter charges for every started thread, regardless of how many agents handle it or how long it lasts. This model aligns the bill with traffic volume rather than team size, and it tends to favour smaller teams on high-traffic sites.

Tidio

Tidio data from its own pricing page, checked September 2026. The Growth tier specifies 250+ conversations; the rate for exceeding that allowance is not published on the self-serve pricing page. The Asyntai row is ours.

Tidio prices per conversation. The free tier allows 50 conversations per month, which is enough to evaluate the product on a very quiet site but not enough to test it at realistic traffic. The published tiers stop at 250 conversations on Growth, above which the price is not self-serve and requires a sales conversation.

The comparison with Asyntai is not straightforward because Tidio counts conversations and Asyntai counts messages. A conversation is a thread; a message is an individual exchange within that thread. A typical support conversation runs four to eight messages: an opening greeting, the question, a clarification, the answer, a confirmation. Some run longer. To make the comparison meaningful, use a per-conversation estimate of five messages as the working assumption.

At the same five-messages-per-conversation estimate, Tidio's Growth tier covers roughly 1,250 message-equivalents. Asyntai's Starter plan covers 2,500 messages. The comparison depends on your actual average messages per conversation: a site with very short interactions might get more from Tidio, and a site with longer conversations might find the Asyntai allowance goes further.

The three pricing models side by side

The groupings above are not marketing categories. They describe what your bill varies with when something changes on your site. This is the practical difference between them:

• Priced per agent or per seat. Your bill changes when you hire or restructure. Adding a weekend shift means adding seats. It does not change when your traffic doubles, unless the extra traffic increases handling time enough to require another person.
• Priced per workspace. Your bill changes when you move to a different tier, typically triggered by seat count or contact volume. It is less sensitive to conversation volume within a tier but becomes sensitive at tier boundaries.
• Priced per conversation or per message. Your bill changes with traffic. A site with 500 conversations in January and 2,000 in February sees a bill that reflects that change, for better or worse.

The model that is right for you depends on what is actually variable in your operation. A staffed inbox with fixed opening hours and a known team size pays a predictable flat fee with per-agent pricing. A site with highly variable traffic pays more predictably with a per-message or per-conversation model, because the bill moves with the workload rather than the headcount.

What a real site pays on each model

To make the comparison concrete, consider a mid-sized e-commerce site that receives 600 conversations per month, runs a chat team of four agents during UK business hours, and uses basic intelligent routing to send returning customers to their previous agent.

For the Tidio and Asyntai comparison, apply the five-messages-per-conversation assumption: 600 conversations is approximately 3,000 messages. Asyntai Standard at 15,000 messages covers this comfortably. Tidio Growth covers 250 conversations; above that, the price requires a sales inquiry.

The comparison shows the gap clearly. Intercom's base platform on Advanced starts at $340 per month for four seats, and the Fin AI add-on scales with whatever the AI resolves. On a busy site, the usage charge can approach or exceed the base platform cost, because you are paying per resolved outcome in addition to per seat. The bill scales directly with AI resolution volume, which on a high-traffic site can be the majority of conversations.

LiveChat and Crisp both come in below Asyntai for this site profile, because four agents put this site on the Asyntai Pro plan. If the site's conversation volume then grew to 5,000 a month, which is 25,000 messages, the LiveChat Team bill stays at $196 and the Asyntai bill stays at $449, because Pro already covers both the seats and the volume. LiveChat's price is set by headcount. Asyntai's is set by whichever threshold you cross first, the seats or the messages.

What changes the answer

The examples above use a specific site profile. Here is what moves the comparison for different types of site.

Seasonal traffic spikes

A retail site that does most of its annual revenue in November and December has a conversation volume that is ten times higher in November than in August. With per-conversation or per-message pricing, the November bill reflects November traffic. With per-agent pricing, the November bill reflects November staffing, which typically scales up to handle the spike. The two models produce similar costs in a high-spike environment, but the per-agent model requires you to staff for the peak, while the per-message model requires you to absorb the peak cost.

High-intent traffic with long conversations

A SaaS site whose visitors are evaluating a complex product often has longer conversations than a retail site. A single evaluation conversation might span twenty messages over an hour. On a per-message model, that conversation uses twenty message allowance units. On a per-agent model, one agent handling it for an hour costs the same as one agent handling ten short conversations in that hour. For high-intent, high-complexity traffic, per-agent pricing can appear cheaper per conversation even though the per-message rate is fixed.

Outside-hours coverage

Live chat that runs 24 hours requires staffing for three shifts in most timezones, or it requires AI coverage outside business hours. Olark's AI Website Agent at $400 per month covers outside-hours with AI rather than staffing a night shift. LiveChat's Business tier can be configured to show an offline form outside hours, which does not generate a per-agent cost but also does not generate an answer.

Asyntai answers at any hour without a separate AI add-on charge, because the message allowance covers AI responses the same as human responses. The cost is the plan price, and the plan price does not change with the time of day.

How to implement each one

Implementation varies by product and by the platform your site runs on. The main methods in common use are a script tag added to the site header, a CMS plugin or integration, and a Code Injection feature on site builders like Squarespace.

LiveChat, Intercom, Olark and Crisp all provide a script tag that you add to the head section of your site. The tag initialises the chat widget and connects it to your account. Most CMS platforms also have official plugins that handle this without editing the theme directly.

Asyntai implements via script tag or via Code Injection on Squarespace, which is Squarespace's built-in method for adding third-party scripts without touching the theme code. The Code Injection feature is available on the Business and Commerce plans.

The routing configuration differs between products. LiveChat's intelligent routing lives on the Business tier and uses rules you define in the dashboard: if a visitor meets condition A, route to team B. Every other vendor here also gates part of its routing behind a tier, and each one draws the line in a different place. Check each vendor's own feature comparison for the tier that carries the rule you need, because the price ladder alone will not tell you.

Two of these products are easy to put live in an afternoon and two are not, and the difference is the routing. A script tag takes minutes on any of them. Writing the rules that decide which team sees which visitor takes as long as your team structure is complicated, so the product that is easy to implement for a two-person team is rarely the best fit for a support desk of twenty.

When you implement any of these chat software examples, plan the routing configuration before you go live, not after. The routing rules determine which team handles which visitor, and retrofitting them once you have live traffic causes more disruption than setting them up correctly from the start.

Sorting the examples by what your bill varies with

Here is the full set, grouped by the meter that drives the actual cost.

Priced per agent or per seat

• LiveChat — per person, billed monthly or annually. Intelligent routing available on Business tier.
• Olark — per agent for live chat; AI Website Agent is a separate flat-rate product.
• Intercom — per seat, with Fin AI as a per-resolved-outcome add-on.

Priced per workspace

• Crisp — per workspace with a block of seats. Extra seats are an add-on on the top tier only.

Priced per conversation or per message

• Tidio — per conversation with a ceiling on self-serve tiers. Above 250, requires a sales conversation.
• Asyntai — per message, with plans from free (100 messages) to Pro (50,000 messages). The message allowance covers both AI and human responses. There is no per-seat multiplier and no per-outcome add-on; each plan carries a block of dashboard seats instead, 2 on Starter, 3 on Standard and 10 on Pro.

The grouping tells you something the individual price ladders do not: which products compete on the same axis. LiveChat and Intercom both pitch per-seat pricing, so a decision between them is a decision about features and support at comparable scale. Tidio and Asyntai both pitch a volume-based meter, so a decision between them is a decision about message allowance and price per message at comparable volume. Crisp sits in its own category, where the workspace is the unit and seat blocks are included rather than priced individually.

What to measure before you pick

Before comparing price ladders, measure three things on your current site. These numbers decide which ladder is the right ladder.

1. Monthly conversation volume. Count every distinct thread started in a representative month, not a peak month. Multiply by your estimate of messages per conversation to compare against per-message pricing.
2. Monthly active agents. Count everyone who logged in to handle chats in a representative month, even once. This is the number that drives per-seat pricing. A site with 12 registered agents who work in rotation pays for 12 seats even if only 6 are active on an average day.
3. AI resolution rate. Estimate what share of conversations the AI could resolve without human input. On a product information site, this can be high. On a complex support case, it may be low. Intercom's Fin AI cost scales directly with this number.

Once you have those three numbers, the grouping above tells you which products to compare directly. If your agent count is stable and your conversation volume is variable, the per-seat products and the per-message products are not comparable; they are answers to different questions. Pick the model that matches what is actually variable in your operation, and then compare the tiers within that model.