Skip to content

By SupportHQ Team · November 20, 2026 · Getting started

AI Customer Support With Flat Pricing: No Per-Seat or Per-Resolution Fees (2026)

AI customer support tools with flat or tiered pricing and no per-resolution fee include Chatbase, Crisp, Wonderchat, SiteGPT, Tidio, and SupportHQ. Of these, Wonderchat and SupportHQ also put no limit or fee on team seats, and Crisp includes up to 10 seats on its mid plan. Intercom, Zendesk, and Help Scout charge per seat and per AI resolution.

“Flat” has a catch. Every flat plan caps something: messages, credits, AI conversations, or conversations. This post explains the three pricing models, prices one team on nine tools, and shows which cap each flat plan has. Every competitor price comes from the vendor’s own pricing page, checked November 2026. Prices change, so confirm them before you buy.

The three ways AI support is priced

ModelYou pay forBill goes up whenExamples
Per seat + per resolutionEach teammate, plus each conversation the AI resolvesYou hire, or the AI resolves moreIntercom, Zendesk, Help Scout
Per AI message or creditA plan with a monthly pool of AI replies or creditsCustomers send more messages, or you pick a costlier modelChatbase, Wonderchat, SiteGPT, Crisp
Per conversation tierA plan with a monthly number of conversationsYou get more conversationsTidio (Lyro), SupportHQ

For a broader price table, see how much AI customer support costs.

The difference that matters most: a per-resolution tool charges only when the AI succeeds. A message or credit plan charges for every AI reply, including replies that do not solve anything. A conversation plan charges for each conversation, whatever the AI does in it.

Worked example: a team of 3 with 1,500 conversations a month

These numbers are assumptions. Replace them with your own.

ToolPlanMonthly totalHow we got it
ZendeskSuite Team$1,6353 × $55 + (750 − 15 included) × $2.00 pay-as-you-go
IntercomEssential$829.503 × $29 + 750 Fin outcomes × $0.99
Help ScoutStandard$637.503 × $25 + 750 AI resolutions × $0.75
WonderchatScale$417Basic has 3,000 AI replies. 4,500 needs Scale (10,000).
TidioPlusFrom $300 + usage1,500 Lyro conversations pass the largest self-serve tier (1,000)
CrispEssentialsAbout $145 to $220$95 + AI usage of $0.05 to $0.10 per conversation, minus $25 included
ChatbaseBasic$160$120 for 4,000 credits + one $40 pack of 1,000 credits
SiteGPTGrowth$7910,000 messages and 4 seats. Starter has 1 seat.
SupportHQStarter$291,500 conversations is under 2,000. No seat fees.

All competitor figures checked November 2026. Notes on each line:

What happens when the AI gets better

Raise the AI resolution rate from 50% to 70%, and keep everything else the same.

ToolAt 50% resolvedAt 70% resolvedChange
Zendesk Suite Team$1,635$2,235+$600
Intercom Essential$829.50$1,126.50+$297
Help Scout Standard$637.50$862.50+$225
Message, credit, and conversation plansNo changeNo change$0

Per-resolution pricing makes success cost more. Your team handles 300 fewer conversations, which is real savings in staff time. The software bill still grows with every extra resolution.

The other direction matters too. At a 20% resolution rate, Intercom costs $87 + 300 × $0.99 = $384, and Help Scout costs $75 + 300 × $0.75 = $300. Credit plans charge the same for a weak AI as for a strong one, yet most of them still cost less here. In this example, the seat fees set the floor, not the AI fees. The case for Intercom, Zendesk, or Help Scout is the help desk around the AI, not the AI price. Compare their email ticketing, routing, and reporting with your needs, not only the AI fee.

The catch in each “flat” plan

A flat plan has a cap. Know what it counts and what happens when you reach it.

ToolWhat the cap countsWhat else is capped
ChatbaseMessage credits, by modelMembers (3 on Basic)
WonderchatAI replies, by modelAI agents per plan
SiteGPTMessages, by modelSeats (1 on Starter, 4 on Growth)
CrispAI credits in dollarsSeats included per plan, then extra agents at $10 a month (check your plan)
TidioLyro conversationsHuman-answered conversations, up to 10 seats
SupportHQConversations in the last 30 daysBots (5 on Starter), website widget messages (50,000 in 30 days on Starter), knowledge base storage, scraping limits

Model choice is the hidden lever in message and credit plans. The same plan can hold 4,000 replies or 1,000, depending on the model you select.

SupportHQ’s limits

SupportHQ sells two plans. Both count conversations, not seats or resolutions.

A conversation is one thread with a customer on the website widget, Telegram, or Discord. You can add your whole team at no extra cost. At the conversation cap, the website widget stops accepting messages, and new Telegram and Discord conversations get no answer, until usage over the last 30 days falls below the cap or you move to a bigger plan. There is no free plan, and the 7-day trial needs a card.

What SupportHQ does not do also affects the comparison. It has no email inbox, no phone channel, and no ticketing. If your team works mainly in email tickets, a seat-based help desk such as Help Scout or Zendesk fits better, and the per-seat price buys real features.

How to choose a pricing model

Answer three questions with last month’s numbers.

  1. How many people answer customers? At 1 or 2 people, seat fees are small. At 10 or more, seats dominate the Intercom and Zendesk bills, and per-workspace or seat-free plans pull ahead.
  2. What share will the AI resolve? The higher the share, the faster a per-resolution bill grows. A flat plan costs the same at 20% or 70%.
  3. How long are your conversations? Count AI replies per conversation. If customers send many messages, a per-conversation plan protects you better than a per-message plan. SupportHQ also has a high message cap on the website widget, so check it if your chats run long.

Then run a two-week test on your shortlist with real traffic. Read the usage page at the end of each week, not just the invoice. For the reasoning behind conversation-based pricing, see how AI support pricing models compare.

Try SupportHQ

Launch an AI support agent grounded in your knowledge base. It answers on your site, in Telegram, and in Discord, and hands off to your team when it matters.