In September 2024, a London-based startup called 11x announced a $24 million Series A led by Benchmark. Nine months later, a $50 million Series B led by Andreessen Horowitz followed. The company's entire pitch revolved around AI-powered "digital workers" designed to perform the full role of a sales development representative. These rounds arrived amid a broader surge of venture capital into AI-powered "agents" targeting specific business functions — from customer support to software engineering — with the AI SDR category attracting some of the largest early checks.

Then, on March 24, 2025, TechCrunch reported that 11x had displayed customer logos for companies that were not actual customers. Former and current employees told the publication that early churn ranged between 70% and 80%. In May 2025, founder and CEO Hasan Sukkar stepped down into a non-executive chairman role, as reported by Bloomberg.

This episode brought clarity to the AI sales tools market, shifting attention toward tools that perform narrower functions well. The following roundup highlights the AI SDR and sales tools that have gained genuine traction.

Copilot vs. Autopilot: Classifying These Tools

A sales development representative (SDR) operates at the front of the sales process. Their responsibilities include building lists of companies worth contacting, identifying the right person at each organization, writing initial outreach emails, following up when there is no response, and passing interested prospects along to the sales team. Much of the work centers on research rather than direct conversations with buyers.

An AI SDR is software designed to take over this role. Users provide a description of their ideal customer, connect email accounts, and the AI handles list creation, email drafting, sending, and even responses. However, different tools serve different purposes within the sales funnel. Some specialize in list building without writing anything. Others grade copy quality but never touch the list. While all of these tools carry the AI SDR label, they are fundamentally different products. The most practical classification depends on how much work the software performs versus how much remains in human hands.

The Four Types of AI Sales Tools

Data tools form the foundation. They identify contacts within target companies, fill in missing company or person information, and monitor events worth reaching out about. These tools produce clean, current lists but do not write or send anything. Outreach remains a human task. Clay exemplifies this category.

Coaches focus exclusively on messaging. As a user composes an email, the tool grades it in real time, flagging lines that sound templated and highlighting sections a buyer might skip. No message is sent until a representative clicks the send button. Lavender built its product around this single function.

Hybrids run campaigns but enforce checkpoints. The AI selects targets and drafts messages, but a human reviews and approves everything before it goes out. Daily volume is capped rather than open-ended. Regie.ai and Salesforge are the leading examples.

Autonomous AI SDRs execute the entire loop without human intervention. Users enter their ideal customer profile, connect mailboxes, and the software finds prospects, writes to them, and follows up unsupervised. This was 11x's original proposition. It is also where outcomes carry the greatest variance, because when the writing is poor, no human exists in the chain to catch it before it lands in thousands of inboxes.

The Honest Performance Picture

Vendor case studies rarely disclose who was on the email list. Reaching out to a cold list of people who have never encountered your brand is fundamentally different from emailing someone the week they changed jobs or secured funding. The first type of cold list typically generates reply rates around 1%. The second can reach 5%.

Instantly's 2026 benchmark report found an average reply rate of 3.43% across its platform. A more revealing statistic in the same report: 58% of all replies came from the first email.

AiSDR, which sells one of these tools, analyzed 75 customer rollouts and published findings more useful than a simple win rate. Teams with an existing, functioning outbound process achieved two to three times better results after adding AI. Teams without one saw no improvement. These tools will not deliver intended results if the underlying outbound process is broken.

The Deliverability Constraint No AI Removes

In February 2024, Google and Yahoo implemented new spam email rules that every tool on this list must comply with. Sending more than 5,000 messages per day to Gmail addresses classifies a sender as a bulk sender. Bulk senders must implement SPF, DKIM, and DMARC — the three DNS records that authenticate mail as genuinely originating from the sender's domain. An unsubscribe button with removal processed within two days is mandatory. Spam complaint rates must remain below 0.3%, with Google recommending under 0.1%.

Microsoft began enforcing its own version of these rules in May 2024. Non-compliant mail is now rejected at the server level rather than simply landing in spam folders.

This deliverability constraint is indifferent to how well AI drafts an email. Volume increases compliance risk, compliance risk damages domain reputation, and a burned domain does not recover simply because a vendor upgraded to a better model. Effective tools manage this constraint by distributing sends across multiple mailboxes and domains, enforcing daily volume limits, and warming new domains for weeks before live campaigns.

The Data and Copilot Layer

Clay: The GTM Data Workspace

Clay is an AI sales tool that resembles a spreadsheet but functions as a full research team. Users enter a list of companies, and each row is checked against one data provider after another until one returns the answer. This process, called waterfall enrichment, is the primary reason Clay finds more emails and phone numbers than competing tools. The platform also features an AI agent called Claygent, which can surface information no database has on file — for example, whether a company operates its own warehouse.

Classification: Data layer. Clay now includes a built-in sequencer capable of sending emails, but that is not its core value proposition. Clay feeds outbound efforts rather than running them.

Pricing: The free plan includes 100 data credits and 500 actions per month. Launch costs $185/month ($167 paid annually). Growth is $495/month ($446 annually). Enterprise pricing is custom. Every plan includes unlimited seats, so adding team members does not increase the bill.

Deliverability posture: Indirect, but more significant than commonly expected. Incorrect email addresses bounce, and bounces damage sender reputation. No sending tool can fix a list that was already broken when it arrived. Most teams blame the sequencer when the real problem started one step earlier.

Pros:

• Match rates substantially exceed single-provider tools, as failed lookups roll to the next vendor
• Unlimited seats on every tier means cost scales with usage, not headcount
• Clay's documentation states users are not charged when an enrichment returns nothing

Cons:

• Credit burn during the learning phase is the most common complaint in user reviews
• The learning curve spans weeks, not days
• Top-up credits carry a 30% premium once monthly allocation is exhausted

Best for: Teams whose outbound underperforms because the list is wrong, not because the copy is.

Lavender: The Email Coach

Lavender operates as a sidebar inside Gmail and Outlook, grading emails in real time as users type. It scores from 0 to 100 based on length, tone, reading level, personalization, and clarity of the call to action, then identifies the specific sentences causing problems. The scoring model was trained on real sales emails and their actual reply outcomes, which is why its suggestions differ from generic LLM rewrites.

Classification: Pure copilot — the safest purchase in this roundup. A human presses send on every message, eliminating the risk of autonomous sending.

Pricing: The free tier covers 5 emails per month. Starter is $29/month, Pro is $49/month, and Teams is $69 per user per month. Annual billing reduces costs by approximately 20%. Ora, Lavender's separate AI agent that writes cold emails rather than coaching, costs $500 per agent per month and includes 1,000 sends.

Deliverability posture: Structurally zero risk. Lavender never touches the user's domain, never rotates mailboxes, and never sends anything. Its only effect on deliverability is positive: improved reply rates signal positive engagement to mailbox providers.

Pros:

• The most affordable way to determine whether copy or list quality is the bottleneck
• Works alongside any existing sequencer, including Outreach, Salesloft, Apollo, and Smartlead — the platforms that defined the sales engagement category before AI-native entrants arrived
• Lavender's own study of 231,818 cold emails found that A-graded messages replied at 4.3%, compared to 3.4% for the rest

Cons:

• Email only — no LinkedIn or phone support
• Chrome extension reliability is a recurring issue in reviews
• The widely cited 580% reply-rate lift reflects a single customer result, not a benchmark

Best for: A founder or small team writing their own outbound who wants emails to stop sounding like everyone else's.

Regie.ai: AI Sequencing With Human Reps

Regie.ai operates sequencing agents, a parallel dialer, and content generation within a single platform called RegieOne. The agents source prospects, draft messages, and prioritize which contacts a rep should focus on next. The rep remains in the loop at all times — this is the core design principle.

Classification: Hybrid, weighted toward copilot. The per-seat pricing model is revealing: a product genuinely designed to replace reps would not charge per rep.

Pricing: AI SEP is $180 per user per month with a 10-seat minimum ($21,600 annual floor). Force Multiplier Rep is $499 per user per month with a 5-seat minimum ($29,940 annual floor). Annual contracts only; no free tier. Parallel dialing and mailbox rotation are available as paid add-ons.

Deliverability posture: The strongest in the hybrid category. Force Multiplier includes 10 provisioned mailboxes with two months of warming built into the package — an unusually transparent approach to pricing the ramp period rather than hiding it.

Pros:

• Mailbox provisioning and warming are included in the plan rather than added as a surprise line item
• The largest independent review base among AI-native tools in this roundup, rated 4.6 across 1,486 G2 reviews
• Published per-seat pricing in a category where most competitors refuse to disclose numbers

Cons:

• The 10-seat minimum excludes any team with fewer than 10 SDRs
• Reviewers consistently report that AI-generated copy requires human editing before sending
• Dialer and mailbox add-ons push actual per-rep cost well above the headline figure

Best for: An existing team of five or more reps that needs to move faster, not smaller.

The Sending Infrastructure

Instantly: Volume Cold Email

Instantly is infrastructure, not strategy. Users connect unlimited mailboxes, upload a list, and the platform distributes sends across all of them so no single inbox or domain absorbs the full load. Warmup, email verification, a 450-million-contact database, and AI agents are included. Approximately 50,000 teams use the platform.

Classification: Infrastructure. It automates sending, not judgment.

Pricing: The Outreach Growth plan is $47/month ($37.60 with annual billing), including unlimited email accounts and unlimited warmup. Hypergrowth is $97/month, and Light Speed is $358/month. Lead credits start at a separate $47/month. Bundles combining both start at $94/month.

Deliverability posture: This is the core product, and it is also where criticism concentrates. Instantly's warmup network is a pool of accounts that email each other to build reputation. In 2026, users reported warmup scores above 90 while live campaigns landed in spam 30% to 40% of the time. A Reddit thread alleged the warmup pool itself caused fresh domains to be flagged. Both accounts surfaced through review roundups rather than firsthand reporting, so they should be treated as reported rather than confirmed.

Pros:

• Unlimited mailboxes and warmup at $47/month undercuts all competitors at this volume
• The 2026 benchmark report provides genuinely useful data
• Higher tiers automatically rotate IPs and swap flagged ones

Cons:

• Warmup scores and actual inbox placement can diverge significantly
• Deliverability diagnostics are limited once problems arise
• Support and billing complaints recur across review sites

Best for: Teams that have already validated their offer and need to scale volume without burning domains.

The Autonomous AI SDRs

Salesforge (Agent Frank): Autopilot With an Honest Toggle

Agent Frank prospects from a 500-million-contact search engine, writes personalized emails, sends across rotated inboxes, and handles follow-ups. What distinguishes it is a toggle switch: Auto-Pilot operates fully unsupervised, while Co-Pilot holds every message for human approval before sending. Salesforge ships both modes without framing Co-Pilot as a temporary training phase.

Classification: Autonomous, honestly labeled. Few vendors in this space place the supervision toggle in the product itself rather than leaving it to the sales conversation.

Pricing: Agent Frank is $499/month, billed quarterly or annually, covering 2,000+ contacts per month at 25 cents each. The Salesforge platform starts at approximately $48/month. Email infrastructure through Infraforge adds from $33/month, bringing a realistic all-in figure to roughly $532/month. Activating Frank requires a demo call.

Deliverability posture: The most operationally transparent in the group. Salesforge owns its sending infrastructure through Infraforge and its warmup through Warmforge. The company states plainly that new inboxes need approximately two weeks of warming before Frank sends anything, with full cold-start setup taking two to three weeks.

Pros:

• The Co-Pilot toggle allows users to supervise for a month before deciding on autonomy
• Infrastructure and warmup are managed within the same login as the agent
• Pricing is a fraction of what 11x-class vendors quote for comparable work

Cons:

• Quarterly minimum commitment required for Frank
• Thin independent review base; much available coverage is written by competitors or by Salesforge itself
• Email-centric, with no buying-signal layer to identify prospects actively in market

Best for: A small B2B team that wants autonomous sending but is unwilling to hand over the send button before observing it in action.

AiSDR: Transparent Pricing, Published Benchmarks

AiSDR is a fully autonomous AI SDR capable of running email and LinkedIn outreach independently while assigning a dedicated human GTM engineer to each account. The tool identifies prospective clients by tracking website visits and social engagement.

Classification: Fully autonomous SDR with transparent pricing.

Pricing: Three tiers are available. Solo, designed for a single user with a limit of 200 AI-researched contacts per month, is $250/month. Explore is $900/month with unlimited users, 800 AI-researched contacts per month, 24/7 Slack support, and dedicated GTM engineer onboarding. Scale is $2,500/month with 2,500 AI-researched contacts and additional features including native two-way Salesforce CRM sync.

Deliverability posture: Standard for the category. Domain warmup runs 30 to 60 days before output reaches steady state. Budget for a slow first month.

Pros:

• Every tier is publicly priced, which almost no other vendor in this roundup does
• $250 entry point enables a genuine test without a quarterly commitment
• AiSDR publishes benchmark data that argues against half its own prospects buying

Cons:

• Above Solo, the commitment reaches $2,700 before results are known
• Playbook-driven setup frustrates users who want custom configurations

Best for: A team that wants to try autonomous outbound without a sales call or annual contract.

Artisan (Ava): The Famous One

Ava sources leads from Artisan's own contact database, researches them, and runs outbound across email and LinkedIn. Artisan is a Y Combinator graduate that raised a $25 million Series A led by Glade Brook Capital on April 9, 2025, with Y Combinator, HubSpot Ventures, and Day One Ventures participating.

Classification: Full autonomous BDR.

Pricing: Demo-gated. Artisan has a pricing page, but it routes to a sales representative. Third-party estimates range from a few hundred dollars per month to over $5,000, and several sources publishing those figures sell competing tools. Annual commitment is standard.

Deliverability posture: Artisan sends email, but reviewers consistently report needing separate tooling for warmup and inbox health monitoring. Buyers should confirm what is bundled before signing.

Disclosure: Three items buyers should know, all reported by TechCrunch:

LinkedIn banned Artisan shortly before Christmas 2025. The reason was not spam. LinkedIn objected to Artisan using its name on its website and alleged the company was purchasing data from brokers who had scraped the platform. Artisan removed the references, audited its data partners, and was reinstated after approximately two weeks.

The "Stop Hiring Humans" billboards were deliberate provocation. CEO Carmichael-Jack told TechCrunch in April 2025 that he does not believe AI will replace people and described the campaign as "mostly just for attention." At the time, the company employed 35 people and was hiring 22 more. Artisan retired the slogan in August 2026 and is now hiring its first human BDR.

Artisan ran subway ads using KC Green's "This is fine" cartoon without permission. Green went public on May 3, 2026, and told followers to vandalize the ads. The two sides settled by May 31, with Artisan pulling the ads in New York and San Francisco.

Pros:

• Genuinely large contact database combining email and LinkedIn in one workflow
• Well funded and unlikely to disappear mid-contract
• Faster setup than building the same stack independently

Cons:

• No published price — budgeting requires a sales call
• Data accuracy diminishes noticeably outside North America
• "The emails feel AI-generated" is the most frequently repeated complaint in reviews

Best for: Teams with budget who want a single vendor for the entire outbound motion and can tolerate pricing opacity.

Qualified (Piper): The Inbound Exception

Piper operates on the user's website rather than working from cold lists. It engages visitors who are already browsing, asks the questions a rep typically would, answers queries, and books meetings directly into Salesforce. Every prospect it interacts with has approached the company first.

Classification: Autonomous, but inbound only. Piper sends nothing cold. Anyone who clicked an ad expecting an AI SDR that fills an empty calendar should understand that this is a different type of product.

Pricing: Demo-gated. Third-party Vendr data suggests typical contracts land between $40,000 and $68,000 per year, though that estimate originates from a competitor's analysis. Salesforce is also required.

Deliverability posture: Largely irrelevant. Piper responds to people already on the user's site rather than pushing into cold inboxes, so bulk-sender rules do not apply.

The most significant fact: Salesforce agreed to acquire Qualified on December 17, 2025, closing the deal on April 1, 2026, as confirmed in its own filings. Piper is being integrated into Agentforce, Salesforce's AI agent platform launched in late 2024 to automate tasks across sales, service, and marketing. This signals which segment of the AI SDR category attracted the largest buyer in enterprise software — the inbound side was acquired, while the cold outbound side continues to debate reply rates.

Pros:

• Does not depend on list quality or domain reputation
• Deep Salesforce integration, built native to the platform
• Best-reviewed product on G2 at 4.9 out of 5 across over 1,500 reviews

Cons:

• Enterprise pricing that excludes most startups
• Requires Salesforce, ruling out HubSpot and Pipedrive users

Best for: A Salesforce-using organization with meaningful website traffic and no one responding to it quickly enough.

What the 11x Saga Taught AI SDR Buyers

Three lessons emerge. First, request customer references you can call directly — not logos on a webpage. Second, treat ARR built from pilots with break clauses as what OnlyCFO termed "Experimental Runrate Revenue," because a customer who can leave at month three does not constitute an annual contract. Third, discard any performance figure that arrives without a stated methodology — including those in this article. A reply rate is meaningless without knowing whether the list was cold or signal-triggered, how many messages the denominator contains, and who conducted the counting.

The Legal Basics

Cold email is legal in the United States under CAN-SPAM, which operates on an opt-out rather than opt-in basis. Businesses may email someone who never requested contact, but they cannot hide. Every commercial message must include accurate header information, a subject line matching the content, a valid physical postal address, and a working unsubscribe mechanism. Opt-out requests must be honored within 10 business days.

Penalties are assessed per message. The FTC maximum penalty is $53,088 per violating email, set in January 2025 and unchanged for 2026. A 5,000-address blast with a broken unsubscribe link constitutes 5,000 violations on paper. In practice, nobody pays the theoretical maximum. Verkada settled the largest CAN-SPAM action on record in August 2024 for $2.95 million. Hiring a vendor to send emails on a company's behalf does not transfer liability.

Europe takes the opposite approach. Under GDPR, an email address constitutes personal data, and B2B outreach typically relies on legitimate interest under Article 6(1)(f). This is a valid legal basis, but it requires a documented balancing test: outreach must target individuals whose role genuinely relates to what is being sold, and the sender must stop immediately upon objection. A scraped list of every contactable EU address fails this test before send.

Final Verdict: Buy the Copilot First, Earn the Autopilot

Start where the risk is lowest. At $29 per month, Lavender reveals within two weeks whether copy is the problem. A Clay plan reveals whether the list is. Together, they identify which half of the outbound process is broken — and neither can burn a domain during the investigation.

Autonomy is earned after something already works. AiSDR's own data makes this case more effectively than any external critic: teams with a functioning outbound motion achieved two to three times better results once AI was introduced, while teams without one gained nothing.

For teams moving toward autonomous tools, the vendors that publish prices and ship supervision toggles deserve preference. Salesforge and AiSDR both do. Those requiring a sales call before revealing pricing have already communicated something important.