Relevance AI is a no-code AI workforce platform built by Relevance AI Pty Ltd. Drag-and-drop tools and agents run on Claude Opus 4.7, OpenAI frontier models, or Gemini 3.1 Pro within the same workflow.
Since September 2025, pricing splits into Actions (what the agent does) and Vendor Credits (LLM model costs). That split changes the cost calculus versus flat-token platforms. In 2026 the public pricing page also restructured: Pro, Team, and Enterprise now all route through Talk to Sales rather than self-serve checkout, with Pro showing 2,500 Actions plus $20 Vendor Credits a month and Team showing 7,000 Actions plus $70 Vendor Credits a month.
System Verdict
Pick Relevance AI if a non-technical team needs AI agents that make judgment calls, not just shuffle data between apps. The sweet spot is sales qualification, support triage, and lead research. Built-in OCR, classify, and extract tools cover most business agent patterns without code.
Skip it if you need engineering-grade control or predictable unit costs. The Actions plus Vendor Credits split is transparent on paper and opaque in practice. Complex agents with multiple decision branches burn credits faster than flat-rate alternatives. The 2026 move to contact-sales pricing makes procurement comparisons harder for self-serve buyers.
Who pays which tier: Free for proof-of-concept (still surfaced in third-party docs, no longer prominent on the vendor page). Pro for solo operators or tight prototypes (2,500 Actions + $20 Vendor Credits per month, talk to sales for price). Team for small sales or support pods (7,000 Actions + $70 Vendor Credits per month, Lindy reports ~$349/mo on monthly billing). Enterprise for custom models, SSO, and contractual SLAs.
Key Facts
| Category | No-code AI workforce builder |
| Launch | 2023-06 |
| Pricing model | tokens), split since Sept 2025; contact-sales-only as of 2026 |
| Free tier | 200 actions/mo · $2 bonus Vendor Credits (per docs; not prominent on the May 2026 pricing page) |
| Pro tier | 2,500 Actions/mo + $20 Vendor Credits/mo · contact sales for price |
| Team tier | 7,000 Actions/mo + $70 Vendor Credits/mo · contact sales · Lindy reports ~$349/mo on monthly billing |
| Enterprise | Custom Actions, Vendor Credits, seats, custom models, SSO, dedicated support |
| Model support | Claude Opus 4.7 · OpenAI frontier models · Gemini 3.1 Pro · multi-model routing |
| Built-in tools | OCR · Classify · Extract · web scrape · PDF parse · email · calling and meeting agents (Team tier) |
| Procurement note | All paid tiers behind Talk to Sales as of May 2026; prior $19 Pro and $234 Team annual rates no longer publicly surfaced |
Every data point above was re-verified against vendor sources on 2026-05-13. See Sources.
What it actually is
A visual builder for composable AI tools and agents. A “tool” is an atomic action like Extract, Classify, or OCR. An “agent” wires tools into a goal-seeking loop with branching, memory, and RAG.
Agents deploy as chatbot widgets, REST endpoints, or scheduled background jobs. Multi-model routing lets one agent call Claude for reasoning, GPT for generation, and Gemini for multimodal in the same run.
The moat is the no-code UX and the opinionated agent primitives, not the underlying LLMs. Anyone with an API key can replicate the tools. Relevance AI bets non-technical teams pay for the integration cut.
When to pick Relevance AI
- Sales ops without engineers. Lead scoring, qualification, and outbound research agents ship in hours.
- Support triage with judgment. Classify incoming tickets, draft responses, escalate when confidence is low.
- Agency client work. Ship AI tools for a client without standing up infrastructure.
- Workforce agents that handle documents. OCR plus Extract plus Classify is the backbone of invoice, contract, and resume workflows.
- Mixed model routing. Single agent that calls Claude for analysis and OpenAI frontier models for drafting without custom plumbing.
When to pick something else
- Developer-grade agent framework: CrewAI or Letta for Python-native control.
- General app-to-app automation with huge SaaS catalog: Zapier or n8n.
- Voice-first conversational agents: Voiceflow.
- Visual LangChain-style agent flows: Langflow.
- Predictable per-seat pricing: Zapier tiers are easier to model than split credits.
Pricing
The public pricing page now routes every paid tier through Talk to Sales rather than self-serve checkout. Per-month allowances are listed; dollar amounts are not.
| Plan | Listed price | Actions | Vendor Credits | Users / scope |
|---|---|---|---|---|
| Free | $0/mo (per third-party docs) | 200/mo | $2 bonus | 1 user, 1 project |
| Pro | Contact sales | 2,500/mo | $20/mo | Solo build user |
| Team | Contact sales · Lindy reports $349/mo monthly | 7,000/mo | $70/mo | Multiple build users + shared projects + calling/meeting agents |
| Enterprise | Custom | Custom | Custom | Unlimited seats, dedicated account manager |
Prices and allowances verified 2026-05-13 via Relevance AI pricing and Lindy’s Relevance AI pricing analysis. The earlier $19/mo Pro annual and $234/mo Team annual rates surfaced through 2026-04-17 are no longer publicly listed; assume budget targets need a sales conversation. All plans include unlimited agents and tools, plus 2,000+ integrations.
Actions are agent steps. Vendor Credits are consumed by the LLM call itself. Complex branching agents burn both faster than simple linear chains. Paid plans support bring-your-own-API-key to bypass Vendor Credits entirely if the customer prefers direct LLM billing.
Against the alternatives
| Relevance AI | CrewAI | Zapier | |
|---|---|---|---|
| Audience | Non-technical ops/sales | Python developers | Mixed, app-automation-first |
| Abstraction | Visual tools + agents | Role-based crews | Triggers + Zaps |
| Agents that reason | Yes, native | Yes, code-defined | Limited via AI Actions |
| SaaS integrations | Narrow | None (you build) | Broadest in the category |
| Cost predictability | Low (split credits) | Medium (LLM tokens only) | High (per-task) |
| Enterprise controls | Yes at Enterprise tier | Yes at Ultra tier | Yes at Enterprise |
| Best viewed as | AI workforce builder | Developer agent framework | App automation platform |
Failure modes
- Pricing moved to contact-sales. As of 2026, every paid tier on the public pricing page routes through Talk to Sales. Self-serve buyers cannot benchmark Pro or Team against alternatives without a discovery call, and prior public rates are no longer surfaced.
- Credit system is opaque. Predicting per-run cost needs manual calibration. Multi-branch agents with retries can spike usage by 5x versus happy-path tests.
- Narrow SaaS integration library. API calls work, but there is no Zapier-scale app catalog. Integrating with a niche CRM means writing the REST call yourself.
- Agent quality depends on prompt craft. The visual builder lowers the floor, not the ceiling. Bad prompts still ship bad agents.
- Production reliability is a DIY concern. Agents drift, loop, or fail silently without human-in-the-loop. Enterprise tier adds audit, not guaranteed determinism.
- Smaller ecosystem. Fewer templates, fewer tutorials, thinner community than CrewAI, LangChain, or Zapier.
- Free tier is demo-sized. 200 actions a month is enough to test, not enough to run a real workflow.
Methodology
This page was produced by the aipedia.wiki editorial pipeline, an automated system that ingests vendor documentation, verifies pricing and model details against primary sources, and generates the editorial analysis you are reading. No individual human wrote this review. Scoring follows the four-dimension rubric at /about/scoring/ (Utility × Value × Moat × Longevity, unweighted average). Last verified 2026-05-13 against Relevance AI pricing, Relevance AI documentation, and Lindy’s Relevance AI pricing breakdown.
FAQ
Is Relevance AI free? A Free tier still exists in third-party docs and references with 200 actions per month plus $2 in Vendor Credits, suitable for demos and early prototypes. The public pricing page in May 2026 foregrounds Pro, Team, and Enterprise behind Talk to Sales; expect to verify free-tier availability directly with Relevance AI before relying on it.
How does Relevance AI pricing work? Since September 2025, pricing splits into Actions (agent steps) and Vendor Credits (LLM token costs). Paid plans also support bring-your-own-API-key to bypass Vendor Credits. As of 2026, Pro lists 2,500 Actions plus $20 Vendor Credits per month and Team lists 7,000 Actions plus $70 Vendor Credits per month, both behind contact-sales. Lindy reports Team monthly billing at $349. Prior public $19/mo Pro and $234/mo Team annual rates are no longer surfaced on the vendor page.
Relevance AI vs CrewAI? Relevance AI targets non-technical teams through a drag-and-drop builder. CrewAI targets Python developers through a role-based framework. Pick Relevance AI when business users own the agent; pick CrewAI when engineers own it.
Does Relevance AI handle OCR and document extraction? Yes. OCR, Classify, and Extract are built-in primitives. Combining them with a knowledge base RAG store covers most document-intelligence workflows without code.
Can Relevance AI use multiple LLMs in one agent? Yes. Multi-model routing lets a single agent call Claude Opus 4.7 for reasoning, OpenAI frontier models for generation, and Gemini 3.1 Pro for multimodal tasks within the same workflow.
Sources
- Relevance AI pricing page: Current tiers, Actions, Vendor Credits
- Relevance AI docs: plans and credits: Action and credit definitions
- Lindy: Relevance AI pricing 2026: Independent pricing breakdown
- G2: Relevance AI pricing: Third-party verification
Related
- Category: AI Automation
