Relevance AI
Platform for building and deploying a workforce of AI agents with a no-code tool builder.
Relevance AI is a no code platform for building what it calls an AI workforce, teams of AI agents that carry out multi step business tasks on their own. It is aimed primarily at the people closest to the work, ops, sales, and support teams and subject matter experts, who can design and deploy agents without relying on developers. A Relevance agent researches, decides, takes actions in other systems, and remembers what it has done.
Building starts in a few ways: you can describe what you want in plain language and have the platform generate a first-draft agent, clone a pre-built one from a marketplace of agents and tools, or assemble one from scratch. Agents are then equipped through a no code tool builder with concrete actions, sending email, updating a CRM, searching the web, or calling any API, and grounded with knowledge by uploading files or syncing from sources like Google Drive, SharePoint, and Notion.
A Workforce coordinates several agents. On a visual canvas, you link specialized agents so each owns a step and hands context to the next. In a prospecting workforce, one agent researches a prospect, a second drafts personalized outreach and a third checks it for quality and tone, all triggered by events in your pipeline or on a schedule. Sensitive steps can be set to wait for a person's approval. Agents keep long term memory across conversations, and on the Enterprise plan evaluations simulate users, score runs against reusable checks, and can block publishing until a version passes.
Relevance AI is model-agnostic, so agents can use whichever language model best balances quality and cost, with a backup model from another provider, and you can bring your own keys. Organizations pick a US, EU or Australian hosting region at signup. The Enterprise plan adds role-based access, audit logs, single sign-on, and trace export, and the company states SOC 2 Type II and GDPR compliance and that customer data is not used to train models. Relevance AI focuses on go to market work, automating research, qualification, outreach, and support.
Vendor details
Canonical URL
https://relevanceai.com
Category
Multi-agent platform
Company status
independent
Use cases & customers
Target customers
Deployment options
In practice
Your prospecting is a manual assembly line: research a lead, write outreach, check it, update the CRM. Relevance AI lets you build a Workforce of specialized agents that each own a step and hand off context automatically.
Your ops team wants to automate work but has no developers. Relevance AI's no-code builder lets subject-matter experts describe an agent in plain language, equip it with actions like CRM updates, and deploy it without code.
You need agents to run against a whole list, not one chat at a time, and stay accurate. Relevance AI runs agents in bulk, grounds them in your synced documents, and holds output to standards through built-in evals.
Sources & related URLs
Agentic Index coverage score
13.0 / 14 capabilities · 93%
| Integrations & Tool Calling | Full |
|---|---|
|
Agents call tools built in a no code tool builder and connect to a catalog of more than 2,000 integrations, and an MCP client lets them use tools from outside MCP servers. Sourcerelevanceai.com/pricingread 2026-09-27 |
|
| Workflow Orchestration | Full |
|
Workforces connect several agents and tools on a canvas with conditions on the edges between them, agents can hand work to subagents, and the tool builder supports looping and branching steps. Sourcerelevanceai.com/docs/build/workforces/workforce-features/approvals-and-escalations.mdread 2026-09-27 |
|
| Knowledge Grounding & RAG | Full |
|
Knowledge sets are built from uploads (CSV, PDF, Excel, JSON, audio), website extraction, manual entry, and syncs from Google Drive, SharePoint and Notion, and an agent either takes the whole set in its prompt or searches it by retrieval with relevance thresholds. Sourcerelevanceai.com/docs/build/knowledge/create-knowledge.mdread 2026-09-27 |
|
| Human Oversight & Guardrails | Full |
|
Approval settings are configured per edge, so a sensitive tool can require approval while routine ones run automatically; in approval required mode the agent drafts its action and waits, and a reviewer approves, rejects or adds guidance from the task view. A let agent decide mode and escalations also exist. Sourcerelevanceai.com/docs/build/workforces/workforce-features/approvals-and-escalations.mdread 2026-09-27 |
|
| Security, Identity & Governance | Full |
|
Relevance AI offers role based access control, SSO and SAML, directory sync, audit logs, PII masking and encryption at rest, and states SOC 2 Type II compliance. SSO, RBAC and audit logs sit on the Enterprise plan. A trust center at trust.relevanceai.com carries the report list. Sourcerelevanceai.com/enterpriseread 2026-09-27 |
|
| Observability & Auditability | Full |
|
Execution traces cover agent invocations, individual LLM calls, workforce runs and condition evaluations, and audit logs cover creation, publishing, deletion and permission changes; on the Enterprise plan both stream to the customer's S3 for OTEL backends. A tasks view shows every agent and workforce run with errors and pending approvals, and views of run cost and outcomes sit alongside it. Sourcerelevanceai.com/docs/enterprise/streaming-events.mdread 2026-09-27 |
|
| Memory & State Persistence | Full |
|
Long term memory lets an agent remember information across different conversations through save to memory and delete from memory tools, alongside short term memory that holds extracted fields for one conversation. Sourcerelevanceai.com/docs/build/agents/build-your-agent/memory.mdread 2026-09-27 |
|
| Deployment & Data Residency | Full |
|
An organization picks its region at signup from three named regions, US (N. Virginia), EU (London) and AU (Sydney), and the region cannot be changed afterward without support. The service is hosted SaaS, and Relevance AI does not describe a self hosted or VPC option. Sourcerelevanceai.com/docs/admin/project-management/ids.mdread 2026-09-27 |
|
| Prebuilt Agents, Templates & Packs | Full |
|
A marketplace carries agents and tools from Relevance staff and verified builders, free or paid, that can be cloned into a project and customized, alongside agent and tool templates and curated multi agent systems such as a BDR agent sold through sales. Sourcerelevanceai.com/docs/get-started/marketplace/introduction.mdread 2026-09-27 |
|
| Triggers & Channel Coverage | Full |
|
Scheduled triggers let an agent act on its own on a recurring schedule or send messages at set future times, and agents also start from integration events, custom webhooks, an API trigger and tools used as triggers. Sourcerelevanceai.com/docs/build/agents/build-your-agent/agent-triggers/scheduled-triggers.mdread 2026-09-27 |
|
| Model Flexibility & Routing | Full |
|
Each agent picks its model from several providers, including Anthropic Claude, OpenAI and Gemini, and can set a backup model, recommended from a different provider, that retries a failed task once. LLM integrations cover OpenAI, Gemini, Anthropic, Azure and OpenRouter. Sourcerelevanceai.com/docs/build/agents/build-your-agent/agent-settings/language-model.mdread 2026-09-27 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
A TypeScript SDK, @relevanceai/sdk, runs agents, tasks, messages and workforces with streaming, an API trigger covers other languages, and an MCP server at mcp.relevanceai.com gives MCP clients OAuth access to a project's agents, tools and knowledge. Sourcerelevanceai.com/docs/sdk/introduction.mdread 2026-09-27 |
|
| Testing, Debugging & Optimization | Full |
|
Test sets hold tests that simulate users and are scored by reusable checks, so versions can be compared on the same tests; publishing can require chosen test sets to pass at a minimum rate and block on failure, and monitor dashboards score live tasks against the same criteria with sampling and trend charts. Agent evaluations sit on the Enterprise plan. Sourcerelevanceai.com/docs/build/agents/evals/introduction.mdread 2026-09-27 |
|
| Browser & Computer Use | Not documented |
|
Browser steps (click, type, scroll by natural language or selector) run through Airtop, a separately sold browser engine that needs the customer's own Airtop account and API key. Relevance AI does not describe a browser or computer use agent of its own. Sourcerelevanceai.com/docs/build/tools/tool-steps/airtop-browser-automation/page-interaction.mdread 2026-09-27 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Product updates
Relevance AI rolled out a multi-feature update introducing the Invent interface for agent creation and adding Google's Gemini 3.5 Flash as a selectable language model across all workflows. The release also updated the platform's MCP server integration to enforce user-level authentication for agent access.
Bears on: Agent capability
View sourceRelevance AI added Confluence Knowledge Sync, project-level public-agent controls, and an Android app, improving enterprise knowledge connectivity, governance, and mobile workforce access.
Bears on: Integrations
View sourceRelevance AI split pricing into Actions and Vendor Credits, enabled BYO API keys, removed markup on vendor costs, and sunset the Business plan.
Bears on: Pricing / packaging
View sourcePricing
Pro at $19 a month billed annually or $29 monthly, Team at $234 a month billed annually or $349 monthly, and Enterprise at custom pricing.
A plan subscription with two meters. Actions count each tool run, and Vendor Credits pay model and tool costs at provider rates with no markup.
Included quota
Pro includes 2,500 Actions and $20 of Vendor Credits a month (30,000 Actions and $240 of Vendor Credits a year on annual billing).
What is public
Three plans are published in Relevance AI's docs. Pro is $19 a month billed annually or $29 monthly (2,500 Actions and $20 Vendor Credits a month, 2 build users, 1 project). Team is $234 a month billed annually or $349 monthly (7,000 Actions and $70 Vendor Credits a month, 5 build users, 45 end users, 5 shared projects). Enterprise is custom. The Free plan is retired and closed to new signups. The pricing page at relevanceai.com/pricing shows only the Enterprise tier with a talk to sales button, so the self serve prices appear in the docs rather than on that page.
Billing mechanics
An Action is a single run of a tool, counted whether it is a simple task or a many step workflow. Vendor Credits are the model and tool cost, passed through with no markup. Bring your own keys is optional on every plan and skips Vendor Credits while keeping them as a fallback. Top ups cost $80 per 1,000 Actions and $20 per 10,000 Vendor Credits. Purchased Action top ups roll to the next cycle, plan Actions reset at renewal, and all Vendor Credits roll over while subscribed.
Cost watchouts
Two meters run at once, Actions for each tool run and Vendor Credits for model costs at provider rates, so always on agents burn both and cost tracks activity rather than seats. Bringing your own model keys, allowed on any plan, skips Vendor Credits. Action top ups at $80 per 1,000 cost more per Action than the plan rate. Evals, SSO and RBAC require Enterprise.
Variable cost rationale
Two things are metered. Every tool run uses an Action, and model and tool costs draw Vendor Credits at provider rates with no markup, so always on agents burn both and the bill tracks agent activity rather than seats. Use beyond the plan's allowance (2,500 Actions and $20 of credits a month on Pro, 7,000 Actions and $70 on Team) is bought as top ups at $80 per 1,000 Actions and $20 per 10,000 Vendor Credits. Customers who bring their own model keys skip Vendor Credits on any plan and pay the model provider directly. Purchased Actions and all Vendor Credits roll over while the subscription is active, and plan Actions reset at renewal.
Overage / add-ons
Top ups cost $80 per 1,000 Actions and $20 per 10,000 Vendor Credits. Bring your own keys (BYOK) is available on every plan. Purchased Action top ups and all Vendor Credits roll over, while plan Actions reset at renewal.
Sales call required
No, self serve available
Free / trial
No free plan or trial, and the Free plan is retired
Lowest paid plan
Pro at $19 per month billed annually ($29 monthly), with 2,500 Actions and $20 of Vendor Credits a month
Commercial notes
Team adds priority support, calling and meeting agents and the analytics dashboard. Enterprise adds agent evaluations, SSO, RBAC and audit logs, enterprise triggers, a dedicated account manager and custom implementation, so governance and evaluation features sit on the Enterprise tier.
Key ambiguities
The pricing page at relevanceai.com/pricing shows only Enterprise with a sales button, while the docs pricing page lists Pro and Team with prices. Annual and monthly figures and two meters appear side by side.
Cancellation / refund
Cancellation and refund terms are not published. Vendor Credits roll over while subscribed, and plan Actions reset at renewal.
Support SLA / resale
Pro gets the community forum only, and Team gets priority support. Enterprise adds a dedicated account manager and custom implementation.
Related vendors
- CrewAI — Open source multi agent framework (Python, MIT) of role based Crews…
- Agent Zero — Open-source multi-agent framework that gives its agent a full Linux…
- Akka — Enterprise agentic AI platform that builds and runs multi agent…
- Artian — Enterprise multi agent AI platform for financial services, where a…
- Databricks Agent Bricks — Databricks' agent platform, now sold as Agent Bricks, for building,…
- Dataiku — Enterprise AI and machine learning platform with an agentic layer…
Alternatives to Relevance AI
The closest documented capability profiles to Relevance AI among multi-agent platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Databricks Agent Bricks13.0 / 14Matches Relevance AI across all 14 documented capabilities
- Dataiku13.0 / 14Matches Relevance AI across all 14 documented capabilities
- Akka12.5 / 14A lighter documented profile than Relevance AI
- CrewAI13.0 / 14Adds documented Browser & Computer UseRelevance AI vs CrewAI →
- Legion Intelligence11.5 / 14A lighter documented profile than Relevance AI
- Rasa11.5 / 14A lighter documented profile than Relevance AI
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded