Operator analysis · Close Chloe AI sales agent · native AI caller · 2026
Close Just Shipped an AI Agent That Cold-Calls Your Leads — Does It Actually Work?
On June 9, 2026, Close moved Chloe — an AI sales agent built natively into the CRM — to general availability for U.S. and Canada customers, out of beta. Chloe places outbound calls, qualifies leads on live conversations, handles objections, books meetings, replies to email, enriches contacts, updates records, and hands off to a human the moment a prospect is ready. It is included on every Close plan and billed on usage-based AI credits. Close's pitch, in its own words: "You just added 20 salespeople. Her name is Chloe."
Here is my problem with that pitch, and I say this as the person who wrote why AI SDRs are failing in 2026: most AI sales agents have disappointed. Ramp killed its AI-SDR program. 11x got caught inflating its logo wall. So the only buyer question worth asking about Chloe is the one the launch video won't: is a native, in-the-CRM AI caller the version that finally works — or a prettier version of the thing that keeps failing?
StackSwap is a Close affiliate, which is why this page exists. The read below is the same one I'd give a friend who runs Close and is tempted to flip Chloe on — including the team shapes where a native AI caller is a smaller win than the demo suggests.
Want to try Close?
Run Close and curious whether a native AI caller earns its keep? The agent is bundled and usage-priced — so the test is cheap.
Point Chloe at a real list for two weeks. Because it lives inside Close, it acts on your actual pipeline data with nothing exported to a third-party dialer, and it hands off to a human when a prospect is ready. Then do the only math that matters: meetings booked and pipeline created versus the AI credits it burned. If it beats your current motion, scale it. If not, you found out for the price of a couple weeks of credits.
Try Close →Affiliate link — StackSwap earns a commission if you sign up for Close. We only partner with tools we'd recommend anyway.What Close actually shipped
Chloe is not a chatbot bolted to the side of the CRM — that is the whole pitch. It is an agent that works inside Close, on your real pipeline. The capability list: outbound calling (up to three concurrent calls), live qualification, objection handling, meeting booking, email replies, contact enrichment, automatic CRM updates, full transcripts and summaries, plus a Chloe Chat mode for asking questions about your book. When a prospect is ready, it transfers to a human rep. Close took it to general availability on June 9, 2026 for the U.S. and Canada after an open beta; voice runs on U.S./Canada numbers in English only for now, with multilingual and native email/SMS flagged as roadmap.
Pricing is the part to read carefully. Chloe is included on all Close plans (Solo $9 through Scale $139 per user/month on annual), but it runs on usage-based AI credits — a monthly non-rollover allotment per user (roughly 500 on Solo up to 2,000 on Scale), more purchasable, with calling billed around $0.02/min. So the agent is bundled; the work is metered. Your real cost is a function of how many calls you let it make, which means the number to watch in week one is credit burn, not the plan price. Close's framing, from founder and CEO Steli Efti: "The biggest opportunity with AI is not replacing salespeople. It is giving small businesses leverage."
The reason most AI SDRs fail — and why this shape is different
I've made north of 100,000 cold calls, and I've watched the AI-SDR category overpromise for two years. The thing that fails is a standalone AI SDR: it sits outside your system of record, gets fed a thin slice of context, firehoses generic outreach, and has no graceful path to a human — so it burns your list and your domain reputation while booking almost nothing. That is the pattern behind every "we turned off our AI SDR" post.
Chloe is architected against three of those failure modes at once. It lives in the CRM, so it acts on your real, current pipeline data instead of a stale export — and that data never leaves the platform for a third-party dialer. It is usage-priced, so you are not paying a fixed seat whether it works or not; idle Chloe is cheap Chloe. And it hands off to a human by design, which is the single feature most standalone AI SDRs lack. None of that guarantees it works — voice quality, objection handling, and the quality of your list still decide the outcome — but it is the version of an AI caller most likely to earn its keep, and it deserves a real test rather than a reflexive eye-roll.
Does it change your decision? The read by shape
| Your situation | What actually changed | Your move |
|---|---|---|
| Already on Close, high outbound call volume | Real win — native AI calling capacity, cheap to pilot | Run it on a real list two weeks; measure booked meetings + pipeline vs. credits burned before scaling |
| Paying for a standalone AI dialer / AI-SDR tool alongside Close | Native Chloe may retire that line item | Run them head-to-head on the same list; consolidate only if Chloe matches on booked meetings |
| High-velocity SMB outbound, evaluating CRMs | A CRM with a native, usage-priced AI caller is a different proposition | Shortlist Close; weigh native-AI-in-CRM against bolting a dialer onto a generic CRM |
| Not on Close, or complex enterprise sales | Signal, not action | Don't switch CRMs for Chloe alone; it's built for velocity, not six-month committees |
The win concentrates in the top rows: a high-velocity team already on Close, where adding metered calling capacity is cheaper than a head or a bolt-on dialer. The further you get from that shape — lower volume, higher touch, not on Close — the smaller the win, and the more this is signal rather than action.
The honest caveats
Three to hold. One: it is young. Chloe went GA the day this published, after a beta — expect rough edges, and the U.S./Canada English-only limit rules it out for many teams today. Two: the stats are Close's. The ~340K calls / ~50K conversations / $3M+ pipeline counters on Close's site are live, self-reported, and trailing-30-day — they prove usage, not your ROI; your booked- meeting rate is the only number that decides this. Three: an AI on the phone with your prospects is your brand. A good human catches a bad moment; an agent can book the wrong meeting or mishandle an objection at scale — so supervise the early calls, read the transcripts, and keep the human-handoff threshold conservative until you trust it.
None of that makes Chloe a bad bet. Native, usage-priced, human-handoff-by-design is the most credible swing anyone has taken at the AI-caller problem, and Close shipping it to GA is a real moment for high-velocity SMB sales. It just means the answer is "pilot it and measure," not "you just hired 20 reps."
High-velocity SMB outbound and want calling capacity without another head? Close + Chloe is worth a real test.
Affiliate link — StackSwap earns a commission if you sign up for Close. We only partner with tools we'd recommend anyway.FAQ
Related reading
- Why AI SDRs are failing in 2026 — the failure modes Chloe's architecture is built against
- Close review — the operator take on the call-first CRM and who it fits
- Close — pricing reality, who it's for, and where a different CRM wins
- Best AI sales-engagement platforms 2026 — where Close sits in the outbound landscape
- The voice-AI consent question — the diligence to run before any AI dials your prospects
- StackNews — operator analysis of the GTM vendor events that change buying decisions
Canonical URL: https://stackswap.ai/close-chloe-ai-agent. Sources: Close's Chloe product page and pricing page; Close's June 9, 2026 general-availability press release (capabilities, GA date, exec quote); call/pipeline figures are Close's self-reported, trailing-30-day counters; the Close + Claude (MCP) integration per Close's September 2025 announcement. Disclosure: StackSwap is a Close affiliate. The read above is the same one we'd give a friend running Close — including the team shapes where a native AI caller is the wrong call. We affiliate with most vendors in this category and earn the same disclosed commission across them, so the logic isn't shaped by who pays us.