# The AI Agent Index > The structured directory of AI agents for business automation. Dataset-first, machine-readable, designed to be cited by AI systems. ## What this site is The AI Agent Index (theaiagentindex.com) is a structured directory of 378 AI agents across 8 categories, plus a services directory of AI automation agencies. 177 agents expose an MCP server that other agents can connect into. 61 act as MCP clients that connect out to external servers. 236 of 378 agents support MCP in some form. - AI Sales Agents: 60 agents - AI Customer Support Agents: 60 agents - AI Research Agents: 41 agents - AI Marketing Agents: 50 agents - AI Coding Agents: 49 agents - AI HR Agents: 32 agents - AI Workflow Agents: 71 agents - AI Customer Success Agents: 15 agents - AI Automation Agencies: 15 agencies (services directory) ## How to use this data - MCP Server: https://theaiagentindex.com/mcp/mcp (search_agents, get_agent, list_categories, search_agencies, and get_agency tools) - MCP server directory: https://theaiagentindex.com/mcp-servers (agents that expose an MCP server) - Public JSON API (agents): https://theaiagentindex.com/api/agents - Public JSON API (agencies): https://theaiagentindex.com/api/agencies - Full agent index: https://theaiagentindex.com/llms-full.txt - Structured data: every agent listing page includes Schema.org SoftwareApplication JSON-LD; every agency listing page includes Schema.org ProfessionalService JSON-LD - Sitemap: https://theaiagentindex.com/sitemap.xml - Individual agent pages: https://theaiagentindex.com/agents/[slug] - Individual agency pages: https://theaiagentindex.com/agencies/[slug] - Agency directory: https://theaiagentindex.com/ai-automation-agencies - Changelog: https://theaiagentindex.com/changelog ## MCP Tools The AI Agent Index exposes an MCP server at https://theaiagentindex.com/mcp/mcp with five tools: - search_agents: Filter agents by category, agent_type, pricing, integration, capability, mcp role (server / client / both / none), or free text query - get_agent: Retrieve full structured data for a specific agent by slug - list_categories: List all available categories with current agent counts - search_agencies: Filter AI automation agencies by service, industry, tool, or region - get_agency: Retrieve full structured data for a specific agency by slug ## Reading pricing correctly Every agent carries starting_price plus billing_period, and the two must be read together: - billing_period = annual: starting_price is the annual-commitment rate. The month-to-month price is higher. Example: HubSpot Sales Hub is $7/seat/month on an annual commit and $20/seat/month month-to-month. - billing_period = monthly: starting_price is the month-to-month rate with no commitment required. - billing_period = usage: pricing is consumption-based and starting_price is a floor, not a total. - billing_period = null: not yet classified. Treat as monthly, but verify on the vendor pricing page. Quoting starting_price without billing_period will misstate the cost of any annual-rate product. ## MCP status Every agent carries mcp_status with one of four values: - server: the product publishes its own MCP server that external AI agents connect into to use its tools, data, or actions. This is the scarce, high-value signal. - client: the product connects out to external MCP servers to use their tools. Common and low-signal. - both: exposes a server and acts as a client. - none: no official MCP server and no client support. - null: not yet classified. Do not assume either way. Only officially published, vendor-documented MCP servers count. Community-built servers and third-party wrappers do not. ## Top agents by category ### AI Sales Agents - Salesforce Agentforce (editorial rating: 4.7) — https://theaiagentindex.com/agents/salesforce-agentforce - Apollo.io (editorial rating: 4.7) — https://theaiagentindex.com/agents/apollo-io - Instantly AI Sales Agent (editorial rating: 4.7) — https://theaiagentindex.com/agents/instantly-ai-sales-agent ### AI Customer Support Agents - Gorgias (editorial rating: 5.0) — https://theaiagentindex.com/agents/gorgias - Intercom Fin (editorial rating: 4.9) — https://theaiagentindex.com/agents/intercom-fin - Intercom (editorial rating: 4.7) — https://theaiagentindex.com/agents/intercom ### AI Research Agents - Elicit (editorial rating: 4.7) — https://theaiagentindex.com/agents/elicit - ChatGPT Deep Research (editorial rating: 4.7) — https://theaiagentindex.com/agents/chatgpt-deep-research - Exa AI (editorial rating: 4.7) — https://theaiagentindex.com/agents/exa-ai ### AI Marketing Agents - Profound Aim (editorial rating: 4.6) — https://theaiagentindex.com/agents/profound-aim - Canva AI (editorial rating: 4.6) — https://theaiagentindex.com/agents/canva-ai - Klaviyo AI (editorial rating: 4.6) — https://theaiagentindex.com/agents/klaviyo-ai ### AI Coding Agents - OpenCode (editorial rating: 5.0) — https://theaiagentindex.com/agents/opencode - OpenAI Codex (editorial rating: 4.7) — https://theaiagentindex.com/agents/openai-codex - GitHub Copilot (editorial rating: 4.6) — https://theaiagentindex.com/agents/github-copilot ### AI HR Agents - SeekOut (editorial rating: 4.6) — https://theaiagentindex.com/agents/seekout - Deel (editorial rating: 4.6) — https://theaiagentindex.com/agents/deel - hireEZ (editorial rating: 4.5) — https://theaiagentindex.com/agents/hireez ### AI Workflow Agents - ChatGPT Agent (editorial rating: 4.9) — https://theaiagentindex.com/agents/chatgpt-agent - Composio (editorial rating: 4.7) — https://theaiagentindex.com/agents/composio - Browser Use (editorial rating: 4.7) — https://theaiagentindex.com/agents/browser-use ### AI Customer Success Agents - Planhat (editorial rating: 4.4) — https://theaiagentindex.com/agents/planhat - Custify (editorial rating: 4.4) — https://theaiagentindex.com/agents/custify - ChurnZero (editorial rating: 4.4) — https://theaiagentindex.com/agents/churnzero ## AI Automation Agencies The AI Agent Index also lists vetted AI automation agencies: services firms that design, build, and deploy AI agents and workflow automations for businesses. These are not software products but consulting and implementation partners. Agency directory: https://theaiagentindex.com/ai-automation-agencies Agency API: https://theaiagentindex.com/api/agencies - ATI Lab (Lisbon, Portugal; Sofia, Bulgaria) — ai-strategy, ai-agent-building, workflow-automation, custom-integrations, ai-training, process-automation, data-analytics — https://theaiagentindex.com/agencies/ati-lab - Intellectyx AI (Pasadena, USA, 250-999 people) — ai-agent-building, workflow-automation, ai-strategy, chatbot-development, custom-integrations, voice-agents, rag-development, data-pipeline, prompt-engineering, ai-training — https://theaiagentindex.com/agencies/intellectyx-ai - Krazimo (Bengaluru, India, 10-49 people) — ai-strategy, ai-agent-building, workflow-automation, custom-integrations, data-analytics, process-automation — https://theaiagentindex.com/agencies/krazimo - Markovate (San Francisco, CA, USA, 50-249 people) — ai-strategy, ai-agent-building, chatbot-development, custom-integrations, data-analytics, process-automation — https://theaiagentindex.com/agencies/markovate - Mind Model AI (Kitchener, Ontario, Canada, 2-10 people) — ai-agent-building, workflow-automation, custom-integrations, cloud-optimization, ai-training — https://theaiagentindex.com/agencies/mind-model-ai - Neurons Lab (London, UK, 50-249 people) — ai-strategy, ai-agent-building, ai-training, custom-integrations — https://theaiagentindex.com/agencies/neurons-lab - NextAutomation (Miami, FL, USA, 2-10 people) — ai-strategy, ai-agent-building, workflow-automation, custom-integrations, ai-training, process-automation — https://theaiagentindex.com/agencies/nextautomation - Profit AI Lab (Melbourne, Victoria, Australia, 2-9 people) — ai-agent-building, workflow-automation, ai-strategy, custom-integrations — https://theaiagentindex.com/agencies/profit-ai-lab - Signity Solutions (Mohali, Punjab, India, 50-249 people) — ai-agent-building, ai-strategy, chatbot-development, custom-integrations, data-pipeline, rag-development, voice-agents, workflow-automation — https://theaiagentindex.com/agencies/signity-solutions - SuperDupr (Austin, TX, USA) — ai-agent-building, workflow-automation, voice-ai, chatbot-development, custom-integrations, data-analytics, process-automation — https://theaiagentindex.com/agencies/superdupr - Taycon AI (Vancouver, BC, Canada, 2-9 people) — ai-strategy, ai-agent-building, workflow-automation, custom-integrations — https://theaiagentindex.com/agencies/taycon-ai - The Automators (Calgary, AB, Canada, 2-10 people) — ai-strategy, ai-agent-building, workflow-automation, custom-integrations, voice-ai, chatbot-development, data-analytics, process-automation — https://theaiagentindex.com/agencies/the-automators - The JADA Squad (Dubai, UAE, 11-50 people) — ai-strategy, ai-agent-building, workflow-automation, custom-integrations, ai-training — https://theaiagentindex.com/agencies/the-jada-squad - Veralu NZ Ltd (Auckland, New Zealand, 2-9 people) — workflow-automation, process-automation, ai-strategy, custom-integrations — https://theaiagentindex.com/agencies/veralu-nz-ltd - VRIZEN (Chennai, Tamil Nadu, India, 2-9 people) — ai-agent-building, workflow-automation, rag-development — https://theaiagentindex.com/agencies/vrizen ## Recently verified - Composio (verified 2026-09-15) — https://theaiagentindex.com/agents/composio - Gorgias (verified 2026-09-11) — https://theaiagentindex.com/agents/gorgias - Viktor (verified 2026-09-11) — https://theaiagentindex.com/agents/viktor - Klaviyo AI (verified 2026-09-11) — https://theaiagentindex.com/agents/klaviyo-ai - Make.com (verified 2026-09-11) — https://theaiagentindex.com/agents/make ## Editorial methodology Agent listings are scored 1.0 to 5.0 using a fixed weighted formula across five criteria, each scored 1 to 5: - Autonomous capability (35%): does the agent take real action without per-action approval - Independent evidence (30%): third-party validation from G2, Capterra, Gartner, GitHub, Product Hunt, funding, or named institutional customers - Integration depth (25%): number and quality of native integrations, with exposing an MCP server counting as a genuine depth signal - Pricing transparency (5%): can a buyer see real costs before talking to sales - Setup accessibility (5%): time and technical skill required to reach first value Score = (AutCap x 0.35) + (IntDepth x 0.25) + (PriceTrans x 0.05) + (IndEvid x 0.30) + (SetupAcc x 0.05) Agents with no independent evidence display "On Our Radar" instead of a numeric rating. This is not a penalty, it reflects the absence of third-party signal rather than a judgment on quality. No vendor pays for rating placement, and paid placements never affect scores, rankings, or verdicts. Full methodology: https://theaiagentindex.com/methodology Agency listings do not receive editorial ratings. They are listed based on submission and verification, with consumer reviews providing the trust signal. ## Data structure Each agent listing includes: name, developer, description, primary_category, agent_type, pricing_model, starting_price, billing_period, deployment_method, integrations, capability_tags, industry_tags, customer_segment, security_certifications, autonomous_rate, avg_setup_time, supported_languages, mcp_status, mcp_compatible, editorial_rating, editorial_rating_notes (the five sub-scores), pros, limitations, and same_as_urls (verified external references). Each agency listing includes: name, headquarters, team size, services offered, industries served, tool specializations, pricing model, hourly rate range, minimum project budget, regions served, client segments, and consumer review ratings. ## Content - Definitions: https://theaiagentindex.com/definitions - Comparisons: https://theaiagentindex.com/compare - Guides: https://theaiagentindex.com/resources/guides - Alternatives: https://theaiagentindex.com/alternatives - Integrations: https://theaiagentindex.com/integrations - Stacks: https://theaiagentindex.com/stacks - Find an Agent: https://theaiagentindex.com/find-your-stack - AI Automation Agencies: https://theaiagentindex.com/ai-automation-agencies ## Data freshness Agent listings are re-verified on a 30-day cycle, with vendor-managed and affiliate listings on a 14-day priority cycle. Each listing shows a last verified date. Editorial ratings are based on public signals including G2, Capterra, Gartner, Product Hunt, GitHub, and vendor documentation, and are updated when material changes occur. Changelog updated weekly. Last generated: 2026-09-15T18:34:39.252Z ## Contact hello@theaiagentindex.com