Questions, answered plainly.

Flatdesk is a help desk for support teams of 3 to 15 agents. Email and website chat land in one inbox, and the AI answers routine questions from your macros. It costs $49 per agent per month, or $39 billed yearly, with AI included and capped.

A ticket is one customer conversation, by email or chat. An agent is a person on your team who answers tickets, and each agent is a paid seat. A macro is a saved reply an agent sends with one click instead of typing it again. Flatdesk's AI writes macros for you from the answers your team keeps retyping.

Yes. When a new email or chat comes in, the AI replies on its own if your facts, macros or help articles cover the question. If they don't, it isn't sure, or the customer asks for a person, it leaves the ticket for your team with a note saying why. An admin can turn this off in Settings.

AI answers reply to customers by themselves, with nobody on your team involved, and each one counts toward the AI answers in your plan. AI macros help your team with the tickets the AI leaves to them: Flatdesk finds replies your team keeps retyping and the AI writes them up as saved replies your team sends with one click. AI macros don't use AI answers.

Email (forward your support address) and a website chat widget you add with one script tag. Both become tickets in the same inbox. Phone, SMS and social channels aren't supported.

Yes, a simple one. Every team gets a public help center with search and writes articles in the app. The AI answers from them and links customers to the right one. There are no categories, languages or custom domains yet, and articles from your old help desk aren't imported.

Run the AI test drive during your trial. After you import your help desk, the AI drafts answers to the 50 most recent tickets your team answered and shows each one beside the reply your team sent. Nothing goes to customers and it doesn't use your AI allowance.

It costs less because of what Flatdesk leaves out, not because the parts you use are worse. AI answers are written by Claude Opus from Anthropic. One answer costs cents to run, so 100 per seat fit in the seat price, and the cap keeps it that way. There's no sales team and only one plan. The product is narrower too: email and chat only, with no phone, SMS, social channels, app marketplace or SLA escalations. Before you pay, the AI test drive shows drafts for the last 50 tickets your team answered, beside its real replies, and the 14-day trial has every feature.

Email and chat teams that want AI to handle a real share of tickets and want the same bill every month. Especially teams tired of paying per AI answer.

Yes, 14 days with every feature and no card, including 25 AI answers per agent for the trial (up to 250 for the team). Add a card any time to get the full 100 per agent a month. The first charge is when the trial ends, or 2 days after you add the card if that's later.

Yes. An admin can make any member who isn't an admin a viewer in Settings. Viewers can read every ticket and report but can't reply or change anything, and they aren't billed.

A conversation the AI finishes without your team. Other help desks call this a resolution. The AI can answer up to 3 follow-ups in the same conversation and it still counts once. If it hands the conversation to your team, including when the customer asks for a person, it doesn't count.

The AI stops for the rest of the month and new conversations go to your team as normal tickets. You aren't charged. You get an email at 80% and at 100%.

Only an admin can turn it on, in Settings under AI answers. Each answer past the included amount then costs $0.40. You can set a monthly limit and turn it off any time.

Per team. A team of 10 agents gets 1,000 a month to share.

No. Monthly billing has no contract. Add or remove seats any time and cancel in Settings. Yearly billing is paid a year ahead for the lower price and runs to the end of that year. If someone leaves mid-year, their seat stays paid until renewal and the next person you add takes it for free.

Yes. Yearly billing is $39 per agent per month, 20% less than $49 month to month, charged as $468 per agent for the year. Same features, same 100 AI answers per agent a month. Seats added during the year are charged for the rest of it. Overage, if you turn it on, is billed monthly.

Yes. Settings exports tickets (with tags), messages, customers and macros as CSV or JSON. Attachments aren't included; download them from each ticket. You never have to ask us.

$49 per agent per month, or $39 billed yearly. Both include every feature and 100 AI answers per agent a month, shared by the team.

No. Some help desks have cheaper starter seats. What Flatdesk does is keep the AI part of the bill from growing on its own. Once the AI answers a real share of your tickets, it usually costs less than tools that charge for every AI answer.

It groups agent replies from the last 90 days by how many words they share. When one answer has gone out on 5 different tickets and no existing macro covers it, it becomes a suggestion, and the AI writes it up.

Flatdesk compares each sent reply with the macro it started from. If agents made the same change on most of the last sends (at least 3), like a new time frame, an added sentence or a removed step, the AI rewrites the macro with that change. You apply it in one click or keep the old one. Zendesk offers macro suggestions, including edits to existing macros, in its Copilot add-on at $50 per agent a month billed yearly. In Flatdesk they're part of every seat.

A macro can assign the ticket to a teammate, set it to open, pending or closed, add tags, and send the reply as soon as it's used. [customer name] is filled in with the customer's first name. A macro set to send right away waits if it still has blanks like [order number], so nothing goes out half done.

No. Writing macros and matching them to tickets never counts toward the AI allowance. Resolutions only count when the AI answers a customer.

The AI writes down the customer question each macro answers. When a new ticket asks the same thing, the reply box offers that macro, and the agent decides whether to use it.

Yes. Imported replies count, so a team that brings its history can get AI macros on the first day.

A monthly statement of every conversation the AI answered, with the macros and articles it used and whether each one counted toward your allowance or overage.

Yes. An admin refunds the line from the receipt. It stops counting toward the allowance, comes off any overage, and the ticket reopens for your team with an internal note.

Until the month closes. Billing runs on the 1st of the next month, and then the receipt is final.

Yes. After you import your help desk, the AI drafts answers to the 50 most recent tickets your team answered and shows each beside the reply your team sent.

No. Drafts stay inside Flatdesk for your team to read and score.

No. It's part of the free trial and doesn't count toward your AI allowance.

Zendesk, Intercom, Freshdesk and Help Scout.

Agents, tags, macros, rules, customers and tickets with their messages and attachments. Intercom's API doesn't share rules. Help Scout workflows come over by name only, since its API doesn't share their conditions. A report lists anything that didn't fit, and every original record is kept.

Yes. Records are matched by their original ID, so running it again updates them instead of making duplicates.

Yes. Settings exports tickets (with tags), messages, customers and macros as CSV or JSON. Attachments aren't included; download them from each ticket.