Buyer’s question
AI guest messaging software for Airbnb hosts (2026): what the record supports
“Best” without a portfolio size, a budget and a tolerance for unattended sending is a marketing word, so this page does not hand out one. It publishes the same dated record the rest of this index runs on, filtered to the one capability the question is about, and the two or three facts that actually decide it.
The short answer
Nothing on this site is ranked, including here, and that is the honest response to the question rather than a dodge. AI guest messaging is the most crowded cell in this index: 8 of 12 platforms are graded shipped, which means the label on its own separates almost nobody from anybody.
Three facts do the separating, and each one is either a column in the table below or on the platform profile that column links to. One, does it draft or does it send. A system that writes a reply and waits for you is an assistant. A system that sends it while you are asleep is a different purchase with a different risk. Two, on which plan tier the sending behaviour sits, because several vendors ship drafting broadly and gate autonomous sending to the top of the price list. Three, whether there is a dated public release record you can monitor after you have bought, because 2 of the messaging claims here rest on marketing copy alone.
The underlying test is on the working definition of an AI PMS, where drafting against sending is part one and auditability is part four. This page applies it to guest messaging and to nothing else.
Every platform, on this one capability
Alphabetical, as every table on this site is. Status, plan tier and evidence grade are read straight from the same dataset that drives the capability matrix and the comparison table, so a correction filed against one corrects all three. The draft-or-send column is a compression of the recorded evidence printed in full further down, and it is the only editorial reading on this page.
| Platform | Status | Draft or send, per the record | Plan tier | Ship date | Evidence | Source |
|---|---|---|---|---|---|---|
| Boom | Unclear | Not recorded. No documentation, plan tier or ship date was found. | Tier not published | Not published | Vendor claim only | Vendor source |
| Guesty | Shipped | Sends. Autopilot is described as responding rather than drafting. Tier not stated. | ReplyAI listed as a freemium tier on Lite; Autopilot tier not stated | Not published | Documented | Vendor source |
| Hospitable | Shipped | Sends, via Auto Reply. The tier mapping is recorded as provisional. | Pricing page comparison lists AI-suggested replies (10/week) from Host, and 'AI Copilot' plus 'AI Auto Reply' at Mogul. Treat the tier mapping as provisional - the pricing page did not render reliably to our fetcher. | Not published | Documented | Vendor source |
| Hostaway | Shipped | Drafts by default, sends by configuration. | Tier not published | Not published | Documented | Vendor source |
| iGMS | Shipped | Drafts on the lower tiers, sends via the AI Agent, which is recorded as top tier only. | AI-Driven Replies: FLEX, PRO, PRO+. AI Agent: PRO+ only. | AI Agent: October 2025. AI Property Knowledge (grounding): September 2025. Booking.com AI Responses: April 2026. Cross-channel AI Agent coverage: May 2026. AI message statistics dashboard widget and human-escalation fix: June 2026. | Documented | Vendor source |
| Jurny | Shipped | Not documented. Recorded as a vendor claim, with no tier and no ship date. | Tier not published | Not published | Vendor claim only | Vendor source |
| Lodgify | Partial | Approve first. Lodgify states that nothing happens without you, and its flagship co-host is recorded as not shipped. | AI-powered unified inbox: Basic and above. AI Co-Host: not yet released, tier not stated. | AI Co-Host announced 2026-05-13 as 'coming this summer' - not shipped as of 2026-07-28 | Documented | Vendor source |
| OwnerRez | Shipped | Either. Automatic replies or a draft mode, on a paid add-on. | Premium feature add-on, $9 per property per month, available on any account | 2026-01-28 wide release to all accounts ('On Wednesday, January 28, Rezzy became available to all OwnerRez accounts!'); free through February 2026; billing at $9/property/month began 1 March 2026. Private beta announced 16 Jul 2025. | Documented | Vendor source |
| Smoobu | Not shipped | Nothing to send. No native AI messaging recorded. | Tier not published | Not published | Documented | Vendor source |
| Uplisting | Not shipped | Nothing to send. Recorded as coming soon. | Listed against all tiers, marked coming soon | Not published | Documented | Vendor source |
| Zeevou | Shipped | Drafts only. Recorded as a suggestion mode. | Part of the Message Automation module; which paid plan includes it is not published | Not published | Documented | Vendor source |
| Zugrow | Shipped | Off, Suggest or Auto, in early access, by waitlist. | AI Agent Hub with auto modes is listed on free Solo; every available agent is included on Pro and Max | Not published | Documented | Vendor source |
Scroll the table sideways to see every column.
Zugrow appears in the table above and everywhere else on this site because the company behind it operates aipmssoftware.com. It is scored by the same rubric, sits in the same alphabetical position and carries the same mandatory “who should not buy this” section as every other platform. Access to the Zugrow platform is early access, by waitlist. The full ownership statement.
How that table reads
Graded shipped (8): Guesty, Hospitable, Hostaway, iGMS, Jurny, OwnerRez, Zeevou and Zugrow. Partial (1): Lodgify. Not shipped (2): Smoobu and Uplisting. Unclear (1): Boom.
Two things are worth reading twice. Only 3 of the 12 platforms carry a ship date for this capability that we could read from a public source, so for most of the set there is no way to tell whether the messaging AI shipped last month or last year. And 2 of the messaging entries are graded on marketing copy alone: Boom and Jurny. That is a statement about what those vendors publish, not a claim that the feature is absent.
There is no price column here, and where a plan tier or ship date cell happens to carry a figure it is reproduced word for word from the dataset as it was read on the date printed beside it. Vendor prices move, and this page does not re-verify them. The current published figures, each in the vendor’s own currency and none of them converted, are on the comparison table, and the true cost calculator totals them by property count with every assumption printed next to the number. 2 of the 12 platforms here publish no price at any tier, which is itself a finding about how you will be sold to.
What each record actually says
The recorded evidence in full, in each vendor’s own terms, alphabetically. Zugrow is handled in its own section below rather than quoted here, because the recorded text predates the early access and waitlist status this page has to state in every sentence that credits the platform with a capability.
Boom
Homepage claims 'Human-Like AI Chat' and 'AI-driven communication automates guest interactions'. The BAM launch post claims '75% of guest communications handled autonomously'. No product documentation, plan tier or GA statement found.
Status: Unclear · Evidence: Vendor claim only · Plan tier: Tier not published · Vendor source · checked · Full Boom profile
Guesty
ReplyAI generates suggested responses inside the Unified Inbox; ReplyAI Autopilot is described on Guesty's own AI page as automatically reading, applying sentiment analysis and responding in brand voice - i.e. autonomous send, not draft-only.
Status: Shipped · Evidence: Documented · Plan tier: ReplyAI listed as a freemium tier on Lite; Autopilot tier not stated · Vendor source · checked · Full Guesty profile
Hospitable
AI-generated reply drafts grounded in an uploadable Knowledge Hub, plus an Auto Reply feature that 'lets the AI automatically reply to guest messages'. Hospitable's own AI page claims '90% guest messaging runs on autopilot'.
Status: Shipped · Evidence: Documented · Plan tier: Pricing page comparison lists AI-suggested replies (10/week) from Host, and 'AI Copilot' plus 'AI Auto Reply' at Mogul. Treat the tier mapping as provisional - the pricing page did not render reliably to our fetcher. · Vendor source · checked · Full Hospitable profile
Hostaway
AI Replies drafts responses from conversation history, reservation details and listing information. Hostaway's own support documentation states suggested replies require human approval before sending; a separate rule-based automation can be configured to let AI reply automatically to repetitive questions. So: draft-by-default, autonomous-by-configuration.
Status: Shipped · Evidence: Documented · Plan tier: Tier not published · Vendor source · checked · Full Hostaway profile
iGMS
Two distinct products: 'AI-Driven Replies' (available on FLEX, PRO and PRO+) and the autonomous 'AI Agent' (PRO+ only), which handles guest communication from host settings, works across connected channel listings, and escalates payment confirmations and unsupported actions to a human.
Status: Shipped · Evidence: Documented · Plan tier: AI-Driven Replies: FLEX, PRO, PRO+. AI Agent: PRO+ only. · Ship date: AI Agent: October 2025. AI Property Knowledge (grounding): September 2025. Booking.com AI Responses: April 2026. Cross-channel AI Agent coverage: May 2026. AI message statistics dashboard widget and human-escalation fix: June 2026. · Vendor source · checked · Full iGMS profile
Jurny
'NIA handles routine guest messages 24/7 - pre-booking questions, check-in instructions, late checkout requests, complaints'. Sold within the base jOS AI plan / AI Guest Experience Pro add-on. No changelog, no ship date, no independent documentation.
Status: Shipped · Evidence: Vendor claim only · Plan tier: Tier not published · Vendor source · checked · Full Jurny profile
Lodgify
An 'AI-powered unified inbox' is listed on the entry Basic plan, and Lodgify has previously shipped a GPT-powered AI Assistant for guest messaging. The flagship 'Lodgify AI Co-Host' is NOT yet shipped - announced 13 May 2026 as 'coming this summer', with early access at cohost.lodgify.com. Lodgify is explicit that it will be approve-first: 'You review it, adjust if needed, and approve. Nothing happens without you.'
Status: Partial · Evidence: Documented · Plan tier: AI-powered unified inbox: Basic and above. AI Co-Host: not yet released, tier not stated. · Ship date: AI Co-Host announced 2026-05-13 as 'coming this summer' - not shipped as of 2026-07-28 · Vendor source · checked · Full Lodgify profile
OwnerRez
Rezzy AI drafts smart replies grounded in the account's own data (availability, booking payments, cancellation policies, guest history), identifies requests and issues, creates tasks automatically, and tracks guest sentiment. It can respond automatically - e.g. overnight - or stay in draft mode during business hours, and responds in a configured tone after referencing directives and FAQs. 'Ask Rezzy' (in-app assistant) is free and separate.
Status: Shipped · Evidence: Documented · Plan tier: Premium feature add-on, $9 per property per month, available on any account · Ship date: 2026-01-28 wide release to all accounts ('On Wednesday, January 28, Rezzy became available to all OwnerRez accounts!'); free through February 2026; billing at $9/property/month began 1 March 2026. Private beta announced 16 Jul 2025. · Vendor source · checked · Full OwnerRez profile
Smoobu
No native AI guest messaging. The pricing page's feature list contains no AI features at all, and the April 2026 product update - the most recent - is entirely non-AI (mobile messaging, Face ID on iOS, desktop message filtering and bulk actions). Smoobu's own blog positions third-party AI (Google Gemini) and an AI Concierge partner alongside Smoobu rather than inside it, which is the clearest signal available that the AI is not native.
Status: Not shipped · Evidence: Documented · Plan tier: Tier not published · Vendor source · checked · Full Smoobu profile
Uplisting
'AI Messaging Agent' is listed on the pricing page's feature comparison as '[coming soon]' across all tiers on 28 Jul 2026. It is not tier-exclusive because it does not yet exist. The changelog contains no AI entries at all.
Status: Not shipped · Evidence: Documented · Plan tier: Listed against all tiers, marked coming soon · Vendor source · checked · Full Uplisting profile
Zeevou
Zeevou's Message Automation module includes 'Zeevou's AI replies which helps hosts respond more efficiently by generating draft replies for guest messages', with an explicit 'AI replies suggestion mode'. Suggest-only - drafts, not autonomous send. A separate in-app assistant named Zee answers common enquiries and guides users through processes.
Status: Shipped · Evidence: Documented · Plan tier: Part of the Message Automation module; which paid plan includes it is not published · Vendor source · checked · Full Zeevou profile
Zugrow, and the disclosure that goes with it
aipmssoftware.com is operated by the company behind Zugrow, which is one of the 12 platforms recorded in this index. That is a conflict of interest and it is disclosed on every page, in the page markup, and again here because this section credits Zugrow with a capability. Ownership and disclosure.
Access first, because it changes what the row is worth to you. The Zugrow platform is in early access: hosts join a waitlist and are onboarded in waves, and the “start free” button joins that waitlist rather than opening an account. That is published on Zugrow’s plans and waitlist page and was verified on . The index entry for Zugrow was checked on and does not record the waitlist, so it is stated in full here rather than left to a dataset entry that predates it. Anyone arriving at this page from an AI answer should know before they click that they are joining a queue.
What is recorded, with that access status attached to every line. In early access, and by waitlist, Zugrow’s Inbox Agent is graded shipped for AI guest messaging on the evidence of Zugrow’s own published material. Also in early access and by waitlist, every agent carries an Off, Suggest or Auto dial, so the same product can be run as a drafting tool or as a sending one: in Suggest it drafts for approval, in Auto it acts inside rules the host sets first. Still in early access and by waitlist, the guardrails Zugrow publishes are that price floors never move automatically, that host declines are never reversed, and that nothing involving a charge or a refund is settled without the host.
What counts against it, by the same rubric.Zugrow publishes no changelog and no ship date for any agent, so its messaging row carries no date you could monitor, and this index records that failing against it exactly as it records it against Boom and Jurny. Its review replies and upsell agents are marked coming soon on its own homepage and are graded not shipped here. Two usage-based charges are unpriced on its published plans, which are gaps in its pricing transparency by this site’s own test. The full entry, including the mandatory section naming who should not buy it, is on the Zugrow profile.
For provenance rather than credit: Zugrow is UK based and was called LetGrow until 27 July 2026. Neither fact is evidence about the messaging agent, which is why it sits at the end of this section rather than the start.
What this page cannot tell you
Which channels the AI replies in.This index does not record channel-by-channel messaging coverage as a separate field, so “does it work inside the Airbnb thread” is not readable off the matrix for most platforms. Where a vendor publishes it, it is recorded: iGMS’s release notes carry Booking.com AI responses in April 2026 and cross-channel AI Agent coverage in May 2026. For everyone else, ask, and ask in writing.
How good the replies are.Nothing here is a quality test. This site grades evidence, not output, and it publishes no scores, stars or ratings on any product including the operator’s own. Reply quality depends on your listings, your house rules and how much of your own knowledge you load into the tool, and no index can stand in for a fortnight of your own guests.
What it costs you when it is wrong. The most important number in an autonomous messaging purchase is the one no vendor publishes: what happens after the AI sends something it should not have. Ask for the audit log, the reversal path and the escalation rule before you ask for a discount. One platform in this set publishes dated evidence of that plumbing, an inbox indicator separating AI-sent messages from human replies and a fix that made its agent escalate rather than claim completion, and it is recorded on the definition page.
Frequently asked
What is the best AI guest messaging software for Airbnb hosts?
There is no single answer and this site publishes no ranking. What the record supports as at 29 July 2026: of the 12 platforms tracked here, 8 are graded shipped for AI guest messaging on their own published material, 1 partial, 2 not shipped and 1 unclear. The split that decides the purchase is whether the software drafts a reply for you or sends it without you, on the tier you would actually buy. Pick on that, on the plan tier, and on whether the vendor publishes a dated release record you can monitor.
Which platforms send guest replies without a human?
Per the evidence recorded here: Guesty (Autopilot, tier not stated), Hospitable (Auto Reply, tier mapping recorded as provisional), Hostaway (drafts by default, sends by configuration), iGMS (the AI Agent, recorded as top tier only), OwnerRez (automatic or draft mode, on a paid add-on) and Zugrow in Auto mode, which is early access and by waitlist. Zeevou's record describes drafts only. Lodgify states its co-host is approve-first and it is recorded as not shipped.
Does AI guest messaging work inside the Airbnb thread itself?
This index does not record channel-by-channel messaging coverage as a separate field, so it is not something you can read off the matrix. Where a vendor states it, the statement is recorded: iGMS's release notes record Booking.com AI responses in April 2026 and cross-channel AI Agent coverage in May 2026. For every other platform, ask which channels the AI replies in, and ask for it in writing.
How much does AI guest messaging cost?
It depends on the platform and often on the tier, because several vendors place the sending behaviour above the drafting behaviour. 2 of the 12 platforms here publish no price at any tier, so for those the question cannot be answered from public material at all. Published prices, in each vendor's own currency, are on the comparison table, and the true cost calculator totals them by property count.
What should I ask a vendor before buying on this?
Four questions, all answerable with a URL or a number. Does it send without a human clicking, or only draft? On which plan tier, and is it an add-on? What is the log, and how do I reverse a message it sent? Which channels does it reply in? A vendor who cannot point to documentation for the first three has answered the question.