Market map

Every listing placed by what it costs and what it asks of you. Nothing is positioned by hand: every cell comes from a field, so the map cannot disagree with the table.

What it costs against what it takes to start

13 of the 39 publish no price at all, so that is a band rather than an omission.

Messaging agents by lowest published monthly price against what a new user must do first
Lowest published price	Text a number	One tap or a group	An account or an app	Gated or self-run	Not recorded
Free to start$0	Arlo Boba Homer	Ember† iChatWithGPT Kachi† Ollie	Muse Poke† Town Vellum Wajo	Hermes OpenClaw OpenInstinct	folk Notis Pally
Under $25	—	Sidekicks	222 Zo Computer	—	—
$25 to $49	—	—	Duckbill Lindy TryOpenClaw	—	Martin
$50 and up	—	—	Catch	—	—
Not publishedthe finding, not a gap	Caddy Ditto Ori Series Tomo	brobot Jarvie Shuffle	Alfi	Boardy Instinct	Lucas Orchid

39 of 39 listings placed, 15 of 25 cells occupied. A dagger marks a listing that is no longer plainly live.

The first column is the one worth staring at. Of the 8 agents you can start using by texting a number, 3 are free and 5 publish no price at all. Not one of them publishes a paid price. The easiest things in this market to try are also the ones that will not tell you what they cost.

What it costs against how much it does

Breadth is what a product claims on its own pages, not what we verified. 18 of the 39 start at nothing; the dearest entry price is $99 a month.

Messaging agents by lowest published monthly price against the number of capabilities they state
Lowest published price	None stated	1 to 3	4 to 6	7 or more
Free to start$0	OpenInstinct	Boba Ember† iChatWithGPT Kachi† Ollie Town	Hermes OpenClaw Poke†	Arlo folk Homer Muse Notis Pally Vellum Wajo
Under $25	—	222 Sidekicks	Zo Computer	—
$25 to $49	—	—	Duckbill Martin TryOpenClaw	Lindy
$50 and up	—	—	—	Catch
Not publishedthe finding, not a gap	—	Alfi Boardy Ditto Jarvie Ori Series Shuffle	brobot Caddy Lucas Tomo	Instinct Orchid

39 of 39 listings placed, 12 of 20 cells occupied. A dagger marks a listing that is no longer plainly live.

Do not read this one as a value map. The banded cells suggest a diagonal and there is no trend underneath it: across every agent that publishes a price, the rank correlation between price and stated breadth is 0.04, which is nothing. The paid bands look like a pattern because they hold one, two and three products. What breadth actually tracks is the job: social agents state 2.5 capabilities on average across 12 listings, against 6.1 across the 24 general assistants. A narrow product is narrow because of what it is for, not because of what it costs.

How to read an empty cell

An empty cell is a finding, not a rendering artifact. Where a whole row or column is thin, that is the market telling you nobody has built the thing yet, or that nobody who has built it is willing to say what it costs.

A dagger marks a listing that is no longer plainly live. Putting the acquired and the beta on the same canvas as everything else is deliberate: a map of only the survivors makes a market look steadier than it is.

Prices are the lowest monthly figure each product publishes on its own pages, last checked 2026-09-09. Where a product publishes only an annual figure, that is the figure used, and the listing carries the tiers in full. Nothing here is estimated. If a product does not publish a price, it sits in the last row rather than being guessed into a band. How we source this.