OutX Overview
A buying-signal product that sends from your own LinkedIn session, which makes the browser on somebody’s desk part of the infrastructure.
OutX is a social-listening product for sales teams. It watches LinkedIn and Reddit for the moment someone posts about a problem you solve, ranks what it finds by intent, drafts a reply, and sends that reply from the operator’s own account through a Chrome extension. The vendor’s one-line version is “LinkedIn social listening that replies.” The extension is the part worth understanding first, because it carries the whole product.
On this page
What OutX is
The homepage states the job in one sentence: “OutX catches the moment a buyer asks for what you sell on LinkedIn or Reddit, and hands you a reply ready to send from your real account.” Everything in the product serves that loop. Keyword watchlists over two networks, a ranked feed of what matched, an AI-drafted reply, and routing into Slack and a CRM so a rep sees it while the post is still warm.
Both networks are live. LinkedIn has been from the start, and Reddit now runs through the official route, which the vendor describes in its own FAQ: “We use Reddit’s official API and your account’s permissions. Comments are sent at human cadence and we surface only public posts.”
The vendor draws its own category line against the nearest adjacent product, and the sentence is useful because it says what OutX thinks it is: “Common Room is a community platform. OutX is a sales tool. We monitor Reddit (they don’t), draft replies (they don’t), and route signals to your AEs in Slack within 30 seconds.”
The outcome numbers on the homepage are the vendor’s own and carry no method: “8 minutes” as the average time to first signal, a “31% connect rate” from signal-led replies, a “3.4x reply rate”, and a single named customer result of “+£480k pipeline / Q1”.
How a signal becomes a reply
The flow has four steps, and the documentation describes each one.
- Watchlists define what to watch. A watchlist holds keywords, profiles and companies, with geo filters attached. Geography is a first-class filter, and the vendor says so: “Geo-routing is first-class. Filter by region, time zone, or language.”
- Scanning happens on a published clock. Data freshness is a plan feature and it changes by tier. The free plan refreshes on a 12 or 24 hour cycle, Growth every 24 hours, Expert every 12, and Ultimate every 6.
- Matches surface ranked by intent, with the post’s context attached, and route onward into Slack, HubSpot, Salesforce or Pipedrive.
- The reply is drafted, then sent as you. A rep clicks to draft a warm reply, and it goes out from their own LinkedIn session. Auto-engagement exists alongside it: auto-likes and auto-comments, where “Comments are AI-generated in a natural tone.”
Capture runs from the same place. Clicking the OutX icon “while on any LinkedIn profile, company page, or Sales Navigator search result” saves the record to a watchlist, with optional email and phone enrichment, and every saved record “pipes into OutX automatically”.
The extension is the runtime
This is the architectural fact that decides how OutX fits a stack, and the documentation states it without hedging. The Chrome extension is not an add-on to the web app. It is where the work happens.
The vendor describes the mechanism as deliberate mimicry: “The extension operates in the background while you browse, it simulates normal LinkedIn browsing, so it looks exactly like you are manually using LinkedIn.” Nothing leaves the browser in the way a cookie-based integration would, and the vendor is explicit about that too: “OutX does not take your cookies, passwords, or save any credentials. Everything runs locally in your browser.” The extension page repeats it from the other side, “we never log your LinkedIn password (we don’t ask for it)”.
Two consequences follow, both documented by the vendor.
- A closed browser stops the product. “If your computer is off or the browser is closed, watchlist scanning and engagement actions will pause and resume when the browser is open again.”
- The API depends on it as well. “The Chrome extension must be installed and active on at least one team member’s browser within the last 48 hours for all features (including the API) to work.”
Read that second line carefully if you are planning to build on the API. A server-side integration written against OutX has an availability condition that lives on somebody’s laptop: a 48-hour heartbeat from at least one team member’s Chrome. For a sales team that opens LinkedIn every working day, that condition is met without anyone thinking about it. For an unattended pipeline that runs at 3 a.m. on a Sunday, it is a dependency that has to be planned around, and a long weekend is enough to break the window.
The extension is under active development. The Chrome Web Store listing shows version 0.0.211, updated on 23 October 2026, published by outx.ai, with a 4.4 rating from 20 ratings and around 6,000 users, while the pricing page prints a “5/5 Chrome Web Store rating”. The install count and the homepage’s “4,000+ sales reps already self-served onto OutX” measure different things, an install base against a signup total, and the extension page adds a third figure, “40k+” active users. All three are worth knowing before reading any of them as the size of the product.
What it costs
Four published plans, priced per month, with a free tier that does real work. Annual billing is marked “SAVE 20%”, and the paid trial terms are stated plainly: “7-day free trial with full access. Card required up front, no charge during the trial, cancel anytime.”
| Plan | Monthly | Leads per week | Profiles tracked | Private watchlists | Data freshness |
|---|---|---|---|---|---|
| Free | $0 | 500 | Limited | 2 | 12 to 24 hrs |
| Growth | $99 | 2,000 | 1,000 | 5 | Every 24 hrs |
| Expert | $249 | 5,000 | 4,000 | 20 | Every 12 hrs |
| Ultimate | $999 | 10,000 | 20,000 | 100 | Every 6 hrs |
Above Ultimate the page offers a custom arrangement covering “White-label options, volume pricing, dedicated account manager, and custom SLAs”. The extension itself has its own free floor, separate from the plan: “The base extension is free with watchlist saving, profile and company tracking, and CSV export”, while “Auto-engagement (auto-likes, auto-comments) and high-volume scraping are gated to paid plans.”
The billing unit is the workspace and its volume, and seats are not the meter. That is a different shape from the per-connected-account pricing the outreach tools in this category use, and it fits the job: one team watching a keyword set does not need an account per rep to do it.
What each plan meters
Three of the four columns above are the real ceilings, and they constrain different things.
- Leads per week is the output cap. It runs 500, 2,000, 5,000 and 10,000, so the top plan is twenty times the free tier and twice the Expert plan at four times the price.
- Profiles tracked is the watch surface. This is the column that moves most between tiers, from 1,000 on Growth to 20,000 on Ultimate, and it is the one to size against a target account list.
- Data freshness is the latency of the whole product. A six-hour cycle on Ultimate against 24 hours on Growth is the difference between replying the same morning and replying the next day, which matters for a product whose pitch is an eight-minute time to first signal.
Sending pace is published as advice, not as a limit. The extension page describes it in words, “OutX uses human-like timing, respects LinkedIn rate limits, and never asks for your password or cookies”, and the API documentation adds a table of what it calls a safe pace per LinkedIn account, among them connection requests “up to 20 a day and 100 a week” and messages “up to 50 a day”, with the note that “The pace below is tentative and OutX does not enforce it.” Because actions run inside the operator’s own browsing session, LinkedIn’s own per-account ceilings are what ultimately apply, and we publish those in full on LinkedIn limits.
What the OutX API covers
The documentation names two API surfaces. A Watchlists & Engagement API covers CRUD on watchlists and engagement automation, and a LinkedIn Data API fetches profiles, posts and connections and also likes and comments on posts, sends messages and sends connection requests. Together they make the product scriptable: a workflow can create a watchlist, read what matched and act on it without a human in the web app.
The reference is public. Calls go to https://api.outx.ai with the key in an x-api-key header, and each endpoint is documented with its method and path. Three things around it are worth knowing. OutX puts no limit on the API itself (“OutX does not enforce API rate limits”), so the pace table above is the caller’s to implement. Webhooks are marked “coming soon”. And the MCP server, 33 tools, runs on your own machine through npx outx-mcp-server, with the docs adding that “There is no hosted or remote endpoint”.
And the API inherits the extension’s condition. The 48-hour rule quoted above says “all features (including the API)”, and that parenthesis is the vendor’s own. An integration built here is a browser-backed integration whatever the transport looks like.
What the vendor says about account safety
OutX answers the ban question directly in its homepage FAQ, and the answer is worth quoting whole because its parts do different work. Asked “Will my LinkedIn account get banned?”, the vendor replies: “No. OutX sends from your real LinkedIn session at human pacing, with no scraping or third-party APIs. Zero accounts flagged across 4,000+ users.”
The architecture behind that sentence is genuine and is the strongest part of the case. No credentials are handed over, no cookies are exported, no server-side session is created, and the actions originate from the same browser the person already uses. On the dimension of where the session lives, that is about as close to manual behaviour as an automated product gets.
The claim’s shape is the part to read carefully, and it is a shape we grade the same way on every vendor in the lane, our own row included. “Zero accounts flagged” is an absolute, and absolutes are the weakest form a safety claim can take, because a single restricted account ends one and LinkedIn restricts accounts for reasons no vendor controls: the member’s own history, the age of the profile, the reputation of the address they connect from. The sample is stated as users, and one user can run several accounts, so it does not convert into an account count. No rate is given that could be argued with.
For the comparison across the whole field, with every vendor’s own wording printed side by side, see the claims table on Best LinkedIn API. On our own stack, gtm-api.com runs 20,000+ accounts with under 1% of them ever restricted, which is a rate with a sample attached.
One more disclosure belongs here. The Chrome Web Store listing states that the extension collects authentication information and location data, and that the developer does not sell it to third parties outside approved use cases. The vendor’s own pages add that “All extension traffic is end-to-end encrypted” and that “OutX is GDPR + CCPA compliant”.
The shape of job it fits
OutX fits a team that already has reps in LinkedIn every day and wants them pointed at the right conversation sooner. The listening layer is the product, the reply draft is the accelerant, and the extension model means nothing about the account changes: the same person, the same browser, the same session, with a queue of warm openings in front of them. For that job the free tier is enough to test the keyword set before any money moves.
The edges are equally clear. The runtime is a browser that has to be open, so coverage is the team’s working hours and a long weekend is a gap. The API is documented as depending on the same heartbeat, which rules out an unattended server-side pipeline as the primary integration path. Volume is capped per plan on leads and tracked profiles. And the safety case is architectural, with an absolute claim over it and no published rate.
Where the job is the other half of the loop, taking LinkedIn actions programmatically on accounts a product owns, with limits enforced before each call and no browser in the path, the category holds account-based APIs: Unipile, Linked API, ConnectSafely.ai and gtm-api.com among them. Signal capture and account-based execution are complements rather than competitors, and plenty of stacks run one of each. The question that separates them is whether the work has to keep going when the laptop closes.
Frequently Asked Questions
What does OutX do?
It watches LinkedIn and Reddit for posts that signal buying intent, ranks the matches, drafts a reply and lets a rep send it from their own account. Keyword watchlists with geo filters define what is watched, matches route into Slack, HubSpot, Salesforce or Pipedrive, and a Chrome extension does the browsing and the sending inside the operator’s existing LinkedIn session. The vendor describes it as LinkedIn social listening that replies.
How much does OutX cost?
Four published plans. Free at $0 with 500 leads a week and two private watchlists, Growth at $99 a month with 2,000 leads a week and 1,000 tracked profiles, Expert at $249 with 5,000 leads and 4,000 profiles, and Ultimate at $999 with 10,000 leads and 20,000 profiles. Annual billing saves 20%, the trial is seven days with a card required up front, and white-label and volume terms are available above Ultimate.
Does OutX need a Chrome extension to work?
Yes. The extension is the runtime, and the documentation says so in as many words. The documentation states that the extension must be installed and active on at least one team member’s browser within the last 48 hours for all features to work, the API included, and that watchlist scanning and engagement pause when the computer is off or the browser is closed. The extension runs in the operator’s own LinkedIn session and takes no cookies, passwords or credentials.
Is OutX safe for a LinkedIn account?
The vendor answers this directly: “No. OutX sends from your real LinkedIn session at human pacing, with no scraping or third-party APIs. Zero accounts flagged across 4,000+ users.” The architecture behind that is real, since nothing leaves the browser and no server-side session is created. The claim itself is an absolute with no published rate and a sample counted in users, and LinkedIn restricts accounts for reasons outside any vendor’s control, so the architecture is the part of the answer worth weighing.
Does OutX have an API?
Yes, two surfaces: a Watchlists and Engagement API and a LinkedIn Data API that reads profiles, posts and connections and can also like, comment, message and send connection requests. Calls go to api.outx.ai with an x-api-key header, OutX enforces no API rate limit of its own, webhooks are marked coming soon, and the MCP server runs locally through npx with no hosted endpoint. The same 48-hour extension condition applies to the API, so a server-side integration still depends on a team member’s browser having been active recently.
Sources & Further Reading
- OutX, homepage (the positioning line, the FAQ answers on bans, Reddit and Common Room, and the outcome figures) · pricing (the four plans, the per-plan ceilings, the annual discount and the “5/5” rating line, with the trial terms in the homepage FAQ)
- OutX, documentation (the 48-hour extension rule including the API, the pause behaviour when the browser is closed, the two API surfaces, and the statement that nothing runs outside the local browser) · API authentication (the base URL, the x-api-key header and the 48-hour rule) · rate limits (no enforced limit, the recommended pace) · MCP (33 tools, local only), read 24 October 2026 · Chrome extension (the free base tier, the paid gates, the capture flow and the pacing statement)
- Google, Chrome Web Store listing (version 0.0.211, updated 23 October 2026, publisher, user count, rating and the data-collection disclosures)
- gtm-api.com, product · pricing (our own numbers)
- Related: Best LinkedIn API · LinkedIn limits · Safe LinkedIn automation · Linked API overview · Unipile overview
LinkedIn actions as typed API calls.
Connect an account server-side, send a connection request or a message as one endpoint, and read the limits before the call goes out. No browser to keep open. On that stack, gtm-api.com runs 20,000+ accounts with under 1% of them ever restricted. Free plan, then from $10 per connected account at volume.
Last updated: October 2026 · Definitions and benchmarks verified at time of publication