What is Relevance AI?
Reviewed Relevance AI's site, docs, and pricing as of 2026-10-26; not deployed yet, so nothing here is hands-on. Relevance AI is a no-code platform for building and recruiting teams of specialist AI agents — sales, customer success, marketing, HR, operations — pitched as an enterprise 'AI workforce' with agent evaluations, tracing, and SOC 2 Type II governance. The catch: its pricing page now lists a single custom 'Talk to sales' plan, while G2 and Capterra still display self-serve tiers (Free, Team $199/mo, Business $599/mo) that the vendor site no longer shows. Review scores split: 4.3/5 on G2 (20 reviews) versus 3.1/5 on Trustpilot from just 2 unanswered reviews.
Key Features
Invent 2.0 — AI solutions engineer
Describe what to automate in plain language; Invent proposes tools, agents, and implementation steps. An embedded deployment team gets your first agent team live in the opening weeks.
Multi-agent workforce builder
No-code specialist agents for sales, customer success, marketing, HR, and operations, mapped to an L1–L4 autonomy framework that runs from assisted copilots to self-driving agents that swap models via evals.
Agent Evaluations, A/B testing & monitoring
Agent evaluations with pass-rate thresholds, A/B testing and analytics, enterprise triggers, real-time monitoring, full agent tracing, and per-task cost visibility with OTEL & Delta Share export.
Enterprise security & integrations
SOC 2 Type II, GDPR, data residency, PII masking, RBAC with SSO/SAML, audit logs, no training on your data; integrations listed as 1,000+ apps on the homepage and 2,000+ on the pricing page.
Pros & Cons
What works
- No-code agent building is the most-praised strength: G2's summary of 20 reviews (4.3/5) puts ease of use and fast setup first, with 13 five-star ratings.
- Invent 2.0 turns a plain-language workflow description into suggested agents and implementation steps — reviewers single it out as the differentiator versus other agent builders.
- Enterprise governance ships built-in rather than as an upsell: SOC 2 Type II, GDPR, data residency, PII masking, SSO/RBAC, audit logs, human-in-the-loop approvals, and a no-training-on-your-data commitment.
- Multi-agent GTM specialization with cost transparency: purpose-built sales agents (Research & Enricher, Meeting Scheduler, Forecast Roll-up, Deal Reviewer) plus per-task cost visibility and OTEL export.
Watch out for
- No public self-serve pricing: as of 2026-10-26 the vendor pricing page shows one Enterprise 'Talk to sales' card, while G2 (Team $199/mo) and Capterra (from $199/mo) listings look stale against a sales-led pivot.
- Hard-line refunds: a November 2025 G2 reviewer was refused a prorated refund after the platform proved unsuitable and was left with 10 months of unused credits.
- Trustpilot sits at 3.1/5 from only 2 reviews in the last 12 months — one citing removed features that broke existing workflows, one describing 'zero customer support' — and the company hasn't replied to negative reviews.
- Credit economics are opaque at the edges: reviewers report burning credits just to clean LLM output between workflow steps, and difficulty predicting credit consumption at scale.
Who is it for?
Mid-market and enterprise GTM/ops teams deploying governed agent workforces
Deployment-team onboarding, evals, SSO/RBAC, and audit logs match exactly what a compliance-sensitive buyer needs before agents touch production data.
Sales teams wanting AI SDR/BDR coverage
Purpose-built sales agents and vendor case studies (Qualified: $7M pipeline across 35+ agents) — all vendor-claimed, so pilot with your own CRM data before an annual commitment.
Agencies and consultants building agents for clients
Invent plus agent templates speed up delivery, and there's an expert partner directory; note reviewer complaints that docs lag behind frequent product updates.
Solo operators and small teams wanting transparent self-serve pricing
No public pricing, a demo-call sales motion, and refused prorated refunds make the smallest budgets the most exposed — simpler builders with published tiers fit better.
Reviewed Relevance AI's site, docs, and pricing as of 2026-10-26; not deployed yet, so this is research from public pages and review platforms, not hands-on testing.
Relevance AI (OnSearch Pty Ltd, Australia) bills itself as the home of the AI Workforce: teams of specialist agents — research, SDR, meeting scheduling, forecasting, contract review — built without code, run on your data, and governed like enterprise software. The standout is Invent 2.0, described on the homepage as your 24/7 AI solutions engineer: you describe a workflow, it proposes the agents and steps. The governance list is long for this category — SOC 2 Type II, GDPR, data residency, PII masking, SSO/RBAC, audit logs, human-in-the-loop approvals, no training on your data — plus agent evals, tracing, and per-task cost visibility.
The Catch
The vendor's own pricing page no longer shows self-serve plans. As of 2026-10-26 it lists one Enterprise card with a Talk to sales button, nothing else. G2 still displays vendor-provided tiers — Free, Team $199/mo, Business $599/mo — and Capterra says from $199/month, so third-party pricing now contradicts the vendor site. If you run a small team, budget a demo call before you can even see a number. Refunds are hard-line: a November 2025 G2 reviewer who found the platform unsuitable was refused a prorated refund and left holding ten months of unused credits. Credit burn is the second theme — one reviewer describes spending credits just to clean up LLM output between steps. Support evidence splits by year: fast, helpful responses fill 2025 G2 reviews, while a January 2026 Trustpilot reviewer describes zero customer support, and the company hasn't replied to either negative review there. A March 2026 review also says a feature the reviewer relied on was removed — product direction is moving fast, in a sales-led direction.
What it actually costs
| Signal | Number shown | Source, 2026-10-26 |
|---|---|---|
| Vendor pricing page | Custom only — 'Talk to sales' | relevanceai.com/pricing |
| G2 (vendor-provided) | Free · Team $199/mo · Business $599/mo | G2 listing |
| Capterra | from $199/mo, flat rate | Capterra listing |
Treat the G2 and Capterra numbers as leads, not quotes. Confirm tier limits, credit allowances, and overage pricing on the call, and get the refund policy in writing before any annual commitment.
Where it fits
Mid-market and enterprise GTM and ops teams get the most out of it — the deployment-team onboarding, evals, and governance assume exactly that buyer. Solo operators who want a published price and a self-serve checkout will be happier with simpler builders like WotNot or BuildMyAgent. If the workload is phone calls rather than CRM and browser work, Synthflow AI is the voice-first counterpart.
The pitch is autonomy-first. The pricing is sales-first.
Relevance AI FAQ
How much does Relevance AI cost?
Does Relevance AI offer refunds?
Is Relevance AI secure enough for enterprise data?
Alternatives to Relevance AI
No-code voice AI agents with in-house telephony and HIPAA-grade compliance — free to build, but the public pricing page now opens at $30,000 a year.
Prompt-to-agent builder priced in credits — cheap entry for AI agencies, opaque per-action billing, and no independent review record yet.