Agentic Index

The best agent infrastructure platforms in 2026, ranked

The best agent infrastructure platforms of 2026 are Agno, Pydantic AI, and Akka, ranked by verified capability coverage on the Agentic Index. All 151 researched vendors in the category are scored against the same 14 capabilities, and Agno documents the broadest coverage at 12.5 out of 14.

Agentic Index ranks all 151 agent infrastructure platforms in its index by verified capability coverage. These are platforms that provide deployment, orchestration, and runtime tooling for production agents. Each vendor is scored across the same 14 point taxonomy used throughout the index, most recently verified on 2026-07-30, and this ranking recomputes automatically as new research is verified.

Buyers also search this category as agent runtimes, agent orchestration infrastructure, and AI agent deployment platforms.

Scored on 14 buyer facing capabilities researched from public sources. No vendor pays for placement. How this ranking works

Best agent infrastructure platforms by buyer need

Best for production reliability

Pydantic AI

Documents 2.0 of 2 on testing and observability, and 12.5 of 14 overall.

Best for developer extensibility

Agno

Documents 2.0 of 2 on APIs, MCP support, and tool calling, and 12.5 of 14 overall.

Best for self hosted and sovereign deployment

Akka

Documents 2.0 of 2 on deployment control and model flexibility, and 12.0 of 14 overall.

Best for complex multi step work

Edge Delta

Documents 3.0 of 3 on orchestration, memory, and knowledge grounding, and 12.0 of 14 overall.

Each segment names the vendor with the most documented coverage on the criteria that segment depends on, computed from the same research as the ranking below. A vendor appears once, in its strongest segment.

1. 1

   A

   Agno12.5 / 14

   High-performance agent runtime and framework (formerly Phidata) with multi-agent support, memory, and Apache 2.0 self-hosting.

   Full coverage on Integrations & Tool Calling and Memory & State Persistence and 9 more.

2. 2

   P

   Pydantic AI12.5 / 14

   Open-source Python agent framework with strong tooling, model flexibility, and observability via Logfire.

3. 3

   A

   Akka12.0 / 14

   Enterprise agentic AI platform that builds and runs multi agent systems on a durable distributed runtime, unifying orchestration, agents, memory, and streaming in one SDK.

   Full coverage on Observability & Auditability and APIs / SDKs / MCP Extensibility and 9 more. No documented coverage on Browser / Computer-use.

4. 4

   E

   Edge Delta12.0 / 14

   Agentic observability platform whose AI Teammates, coordinated by the OnCall AI super agent, autonomously triage alerts, run root cause analysis, review code and PRs, and file tickets across SRE, security, and DevOps work, grounded in real time telemetry.

5. 5

   I

   Infobip12.0 / 14

   Global CPaaS leader whose AgentOS orchestration layer runs autonomous AI agents across more than fifteen channels, with built in MCP servers and human in the loop oversight.

6. 6

   K

   Kestra12.0 / 14

   Open source, event driven orchestration platform that unifies data, AI, infrastructure, and business workflows behind a declarative YAML interface, with AI native agents, human in the loop approvals, and more than one thousand two hundred plugins.

7. 7

   S

   SnapLogic12.0 / 14

   Enterprise iPaaS repositioned as the Agentic Integration Company, letting teams build and run low-code LLM agents on top of thousands of prebuilt connectors, with MCP support and hybrid deployment.

8. 8

   B

   Bluebricks11.5 / 14

   Context and control layer that lets AI agents operate cloud infrastructure safely, grounding them in a live cloud graph and routing every change through governed plan, approve, and apply orchestration.

   Full coverage on Prebuilt Agents / Templates / Packs and Triggers & Channel Coverage and 8 more. No documented coverage on Browser / Computer-use.

9. 9

   C

   Celonis11.5 / 14

   Process intelligence platform whose AgentC suite and Process Intelligence Graph give AI agents the business context to analyze, design, and operate autonomous enterprise processes.

   Full coverage on Workflow Orchestration and Model Flexibility & Routing and 8 more. No documented coverage on Browser / Computer-use.

10. 10

    F

    FLOWX.AI11.5 / 14

    AI native agent platform for banking that deploys 220+ prebuilt agents on top of legacy systems with an agent builder, in perimeter data control, and banking grade governance.

    Full coverage on Human Oversight & Guardrails and Deployment & Data Residency and 8 more. No documented coverage on Browser / Computer-use.

11. 11

    G

    Gnani.ai11.5 / 14

    Voice AI platform with foundational speech models and agents that run production workflows for enterprises.

12. 12

    L

    LangChain11.5 / 14

    Open-source framework for building LLM-powered agents and multi-agent pipelines, with observability via LangSmith and graph orchestration via LangGraph.

    Full coverage on Deployment & Data Residency and Knowledge Grounding & RAG and 8 more. No documented coverage on Browser & Computer Use.

13. 13

    P

    Paragon11.5 / 14

    AI-native integration infrastructure that gives agents a thousand plus prebuilt actions across one hundred thirty plus apps via a single API or MCP server, with triggers, RAG sync, and workflows.

14. 14

    T

    Tray.ai11.5 / 14

    Enterprise automation and AI agent platform with MCP-native integration capabilities.

    Full coverage on Memory & State Persistence and Integrations & Tool Calling and 8 more. No documented coverage on Browser & Computer Use.

15. 15

    D

    Datafold11.0 / 14

    Data engineering automation platform whose AI Migration Agent translates and validates data pipelines to full parity, with a Data Knowledge Graph and MCP tools that make coding agents reliable.

    Full coverage on Testing, Debugging & Optimization and Integrations & Tool Calling and 7 more. No documented coverage on Browser / Computer-use.

16. 16

    I

    Inworld AI Agent Runtime11.0 / 14

    Realtime agent runtime for graph-based AI systems: provides orchestration nodes, telemetry, memory, experiment support, on-prem TTS, and deployment options for AI product teams. Infrastructure layer, not an end-user assistant.

17. 17

    I

    Island11.0 / 14

    The enterprise browser as agent control plane: build and run governed autonomous agents with per agent permissions, human approval gates, real time DLP on MCP data flows, per invocation SIEM audit logs, and model agnostic provider choice.

18. 18

    M

    Mastra11.0 / 14

    TypeScript agent framework with evals, workflow orchestration, observability, and Mastra Studio control plane, with dual licensing.

    Full coverage on Deployment & Data Residency and Browser & Computer Use and 8 more. No documented coverage on Knowledge Grounding & RAG and Human Oversight & Guardrails.

19. 19

    S

    Stacklok11.0 / 14

    Enterprise MCP platform (built on open source ToolHive) that runs, governs, and secures MCP servers and agent skills in isolated containers, Kubernetes native and self hosted, from the Kubernetes co creators.

20. 20

    A

    AutoGPT10.5 / 14

    Open-source AI agent platform with a visual block-based builder, agent marketplace, and continuous agent runtime, available as free self-hosted software or a hosted cloud with subscription plus pay as you go credits.

    Full coverage on APIs / SDKs / MCP Extensibility and Model Flexibility & Routing and 6 more. No documented coverage on Browser / Computer-use.

21. 21

    A

    Avesha10.5 / 14

    Autonomous AI SRE platform whose coordinated agent swarm monitors, diagnoses, and remediates issues across Kubernetes and cloud, with root cause analysis, auto remediation, and GPU orchestration.

    Full coverage on Triggers & Channel Coverage and Memory & State Persistence and 7 more. No documented coverage on Browser / Computer-use and Model Flexibility & Routing.

22. 22

    F

    Forest Admin10.5 / 14

    Ops orchestration layer for fintechs where humans and AI agents run compliance workflows on your own data, MCP native, bring your own LLM, SOC 2 certified with full audit trails.

    Full coverage on Deployment & Data Residency and Security, Identity & Governance and 6 more. No documented coverage on Browser / Computer-use.

23. 23

    S

    Sendbird10.5 / 14

    Omnichannel AI agent platform for customer service, built on Sendbird's communications APIs, with persistent long term memory across chat, voice, SMS, email, and social.

24. 24

    SmythOS10.5 / 14

    Multi-agent operating system for orchestrating and deploying AI agents at scale.

25. 25

    Vapi10.5 / 14

    Voice agent platform for building, testing, and deploying phone and web voice agents: telephony infrastructure, tool calling, real-time monitoring, and self-serve pricing for developers.

26. 26

    xpander.ai10.5 / 14

    Backend as a service and runtime for AI agents, providing memory, a 2,000 plus tool and MCP repository, multi agent orchestration, and governed deployment.

    Full coverage on Integrations & Tool Calling and Workflow Orchestration and 5 more.

27. 27

    Airtop10.0 / 14

    Cloud browsers for AI agents that compile plain English workflows into deterministic agents: runs on schedules or triggers, self-heals broken runs, persists authenticated profiles, and supports Live View human takeover. SDKs, REST APIs, and a builder agent named Mark.

28. 28

    Corti10.0 / 14

    AI infrastructure for healthcare developers: clinical ASR, healthcare LLMs, and an Agentic Framework with orchestration, guardrails, and an Agent Library through one API, with audit trails on every call and usage based credit pricing.

29. 29

    Letta10.0 / 14

    Memory-first stateful agent platform with an open-source runtime, persistent agents, context repositories for knowledge grounding, model-agnostic support, and a cloud-hosted path for AI developers.

30. 30

    Moderne10.0 / 14

    Enterprise platform on the open source OpenRewrite engine that runs deterministic refactoring recipes across thousands of repositories, with the Moddy multi repo AI agent, Prethink codebase context for coding agents, and an MCP server exposing recipes as tool calls.

31. 31

    Netlify10.0 / 14

    AI native web development platform where developers and AI coding agents build and deploy together, whose Agent Runners bring Claude Code, Codex, and Gemini into the dashboard with full production context, sandboxed execution, and human approved previews, alongside an AI Gateway, MCP server, and observability.

32. 32

    Pipecat10.0 / 14

    Open-source framework for building real-time voice and multimodal AI agents.

    Full coverage on APIs, SDKs & MCP Extensibility and Model Flexibility & Routing and 5 more. No documented coverage on Browser & Computer Use.

33. 33

    PRE Security10.0 / 14

    AI native predictive SecOps platform (Predict & Prevent) that replaces or augments legacy SIEM with parserless ingestion, generative XDR, agentic triage, and predictive analytics.

34. 34

    Snowplow10.0 / 14

    Behavioral data platform and customer context layer that collects, validates, enriches, and delivers real time event level data to a company's own warehouse, lake, or stream, grounding AI agents and agentic analytics in governed, high quality first party data.

35. 35

    Thread AI10.0 / 14

    Composable AI orchestration platform whose Lemma product lets enterprises build and run proprietary agents and workflows on an extended serverless workflow engine, with native human in the loop and tool calling, strong governance, and VPC or on premise deployment.

36. 36

    Haystack9.5 / 14

    Open-source LLM orchestration framework by deepset for building RAG, search, and agentic pipelines, with enterprise support.

    Full coverage on Workflow Orchestration and Integrations & Tool Calling and 5 more. No documented coverage on Human Oversight & Guardrails and Browser & Computer Use.

37. 37

    Itential9.5 / 14

    Agentic infrastructure orchestration platform whose MCP Server and FlowAI framework let AI agents act on networks and hybrid cloud through governed workflows with RBAC, approval gates, validation, rollback, and full audit trails.

38. 38

    Linx Security9.5 / 14

    AI native identity security and governance platform that maps and governs human, non-human, and AI agent identities in real time, with an autonomous remediation agent, Autopilot.

39. 39

    nexos.ai9.5 / 14

    Enterprise AI control layer from the Nord Security founders pairing an AI Gateway, one secure endpoint routing more than two hundred models with caching, fallbacks, RAG, and spend controls, with a governed AI Workspace where employees chat and build reusable agents under guardrails and full logging.

40. 40

    Veecle9.5 / 14

    Agent infrastructure that gives AI agents access to virtual chips, hardware knowledge, and physics simulation via a single API and MCP, so they can build, run, and test firmware without hardware.

    Full coverage on Knowledge Grounding & RAG and Integrations & Tool Calling and 5 more. No documented coverage on Model Flexibility & Routing and Browser / Computer-use.

41. 41

    Waniwani9.5 / 14

    Distribution and compliance infrastructure for selling financial services through AI, with a free open source SDK to build agents that quote inside ChatGPT, Claude, and WhatsApp, plus paid compliance and analytics modules.

42. 42

    Apify9.0 / 14

    Cloud platform for web scraping and automation with 45,000+ prebuilt Actors, proxies, and an MCP server that exposes them as tools for AI agents.

    Full coverage on Integrations & Tool Calling and Security, Identity & Governance and 4 more. No documented coverage on Human Oversight & Guardrails and Model Flexibility & Routing.

43. 43

    Chamber9.0 / 14

    AIOps agent for ML teams that puts GPU infrastructure on autopilot: Chambie monitors, root causes and autonomously fixes failed training runs, reruns from checkpoint, and right sizes and schedules GPU workloads across clouds, deployed inside your own cluster.

    Full coverage on Observability & Auditability and APIs / SDKs / MCP Extensibility and 4 more. No documented coverage on Browser / Computer-use and Model Flexibility & Routing.

44. 44

    DBOS9.0 / 14

    Open source durable execution library built on Postgres that makes workflows, queues, and AI agents fault tolerant and observable with a few code annotations, managed centrally through the Conductor workflow ops console.

45. 45

    Fluid Topics9.0 / 14

    Product knowledge platform that unifies technical content into an AI ready hub and, through core RAG services and an MCP server, grounds chatbots, copilots, and agents in trusted, cited product information.

    Full coverage on APIs / SDKs / MCP Extensibility and Human Oversight & Guardrails and 5 more. No documented coverage on Browser / Computer-use and Deployment & Data Residency and 1 more.

46. 46

    Hatchet9.0 / 14

    Durable orchestration engine for AI agents, background tasks, and mission-critical workflows, built on PostgreSQL with automatic retries, checkpointing, and observability, available as managed cloud or MIT-licensed self-host.

    Full coverage on APIs / SDKs / MCP Extensibility and Observability & Auditability and 5 more. No documented coverage on Knowledge Grounding & RAG and Browser / Computer-use and 1 more.

47. 47

    LlamaIndex9.0 / 14

    Data framework for building RAG pipelines and data agents with LlamaParse document ingestion and LlamaCloud managed platform.

    Full coverage on APIs, SDKs & MCP Extensibility and Model Flexibility & Routing and 4 more. No documented coverage on Browser & Computer Use and Triggers & Channel Coverage.

48. 48

    Mindgard9.0 / 14

    Attacker-aligned AI security platform running continuous automated red teaming against complete AI systems: reconnaissance maps the production attack surface (guardrails, system prompts, tools, shadow AI), then thousands of MITRE ATLAS and OWASP-mapped attacks test LLMs, agents, and multimodal models, including chained exploits, with runtime protection in production. Developer-native deployment via GitHub Action, CLI, APIs, and Burp Suite against any inference endpoint. Lancaster University spinout, 11 PhDs, SOC 2 Type II, $11.6M raised, 2025 Cybersecurity Excellence Award winner.

49. 49

    Nektar.ai9.0 / 14

    GTM telemetry and Customer Brain layer that auto captures every customer interaction into CRM, data warehouse, and AI apps to give agents trustworthy revenue data.

    Full coverage on Observability & Auditability and Knowledge Grounding & RAG and 4 more. No documented coverage on Model Flexibility & Routing and Browser / Computer-use.

50. 50

    Operant AI9.0 / 14

    Runtime AI application defense platform protecting every layer of live cloud and AI applications through discovery, detection, and defense, deployed via single-step Helm inside the customer's own Kubernetes cluster with zero instrumentation. Agent Protector (Feb 2026) adds agentic security: prompt-to-tool-to-memory tracing, intent analysis, continuous runtime re-authorization of tool calls, and memory poisoning protection, integrating with LangGraph, CrewAI, n8n, and the ChatGPT Agents SDK. Also ships AI Gatekeeper, MCP Gateway, and open-source Woodpecker red teaming. The only vendor in six Gartner AI security reports.

51. 51

    Phonic9.0 / 14

    Voice AI platform built on proprietary speech-to-speech audio foundation models: build, observe, and evaluate reliable agents with ~300ms latency, an adaptive decision system instead of rigid state machines, a system of record, and cloud or on-prem deployment. Seed-stage, oriented to healthcare and insurance.

52. 52

    Pillar Security9.0 / 14

    Unified AI security platform where offense and defense work as one system across the AI lifecycle: discovery and posture (AI-SPM cataloging models, tools, MCP servers, coding agents), a first-class Gartner-recognized red-teaming pillar (RedGraph offensive agentic testing, tool-use and black-box red teaming), runtime guardrails over agentic tool calls and permissions, and governance and compliance. A red-team finding becomes a policy becomes a retest, so protection compounds. Tops the AI-agent-security cluster on the strength of its offensive-testing loop.

53. 53

    Repello AI9.0 / 14

    Enterprise platform for autonomous AI red teaming, runtime guardrails, and continuous AI security across LLMs, agents, and MCP, run on a three-phase loop: discover and build an AI BOM, red-team, then deploy adaptive guardrails calibrated from results. ARTEMIS profiles and attacks like a human red-teamer, chaining multi-stage exploits across integrations and visualizing attack paths, drawing on 15M+ attack patterns with multi-modal and multi-lingual coverage mapped to OWASP, NIST, and MITRE. Black-box, CI/CD-integrated, 270+ vulnerability types; publishes open-source tooling and original AI-threat research.

54. 54

    Smallest AI9.0 / 14

    Full stack voice AI platform combining Atoms, a no code voice and chat agent builder handling conversations in more than twenty languages, with its own co optimized models: Lightning text to speech at roughly one hundred millisecond latency and Pulse speech to text at sixty four millisecond first transcript.

55. 55

    TrueFoundry9.0 / 14

    Enterprise AI gateway and MCP gateway unifying 250 plus models with routing, guardrails, observability, and governance across SaaS, VPC, and air gapped deployments.

56. 56

    AlphaBitCore8.5 / 14

    AI control plane that governs how models, agents, tools, and workflows execute inside regulated financial firms, with policy enforcement, sealed audit records, and provable replay.

    Full coverage on Workflow Orchestration and Security, Identity & Governance and 5 more. No documented coverage on Browser / Computer-use and Memory & State Persistence and 2 more.

57. 57

    BotGauge8.5 / 14

    Autonomous QA as a service where AI agents generate, execute, and maintain end to end tests, validated by a human expert pod.

    Full coverage on Browser / Computer-use and Security, Identity & Governance and 3 more. No documented coverage on Prebuilt Agents / Templates / Packs and Model Flexibility & Routing.

58. 58

    Clawvisor8.5 / 14

    Open source authorization gateway for AI agents: vaults credentials, enforces task scoped access, and puts a human approval gate and full audit trail on every tool call.

    Full coverage on APIs / SDKs / MCP Extensibility and Model Flexibility & Routing and 5 more. No documented coverage on Testing, Debugging & Optimization and Browser / Computer-use and 2 more.

59. 59

    HiddenLayer8.5 / 14

    Leading independent AI/ML model security platform extended into agents: Model Scanner (35+ formats, supply-chain threats), AI Detection and Response against adversarial attacks and prompt injection, model genealogy and AIBOM governance, plus 2026 agentic runtime security that detects malicious tool calls and data exfiltration in multi-step autonomous workflows. The model-security-first counterpart to agent-native Noma and Zenity.

60. 60

    Lasso Security8.5 / 14

    GenAI-first, end-to-end AI security platform covering models, agents, and applications through a five-pillar framework: Discover (shadow AI and agent inventory), Assess (posture on what agents can reach and do), Test (automated red teaming with a 3,000+ attack library and multi-turn agentic attacks), Enforce (inline policies at the proxy, API, or AI gateway layer), and Protect (runtime detection and response). Its Intent Security framework, anchored by the Intent Deputy launched in 2026, sets behavioral baselines for agentic AI and covers MCP security. Sub-50ms classifications with a published 98.6% accuracy rate.

61. 61

    Metorial8.5 / 14

    Open source MCP integration platform and access control plane that connects any AI model to 1,000+ tools with serverless hosting, unified auth, RBAC, and audit logs.

    Full coverage on Deployment & Data Residency and Prebuilt Agents / Templates / Packs and 4 more. No documented coverage on Browser / Computer-use and Knowledge Grounding & RAG and 1 more.

62. 62

    Noma Security8.5 / 14

    Market-leading independent enterprise AI and agent security platform: discovery and posture (AI-SPM), Agent Access Control (tool-level governance of agents and MCP servers), runtime AI Detection and Response across prompts, tool calls, MCP, and agent-to-agent traffic, plus continuous AI Red Teaming. Covers homegrown, SaaS, and developer-machine agents. The leading independent counterpart to Zenity.

63. 63

    Sales Layer8.5 / 14

    Agentic product information platform whose autonomous agents enrich, translate, optimize, and distribute catalog data, chaining multi step actions with a review gate and an MCP server.

    Full coverage on Knowledge Grounding & RAG and APIs / SDKs / MCP Extensibility and 5 more. No documented coverage on Deployment & Data Residency and Model Flexibility & Routing and 2 more.

64. 64

    Senso8.5 / 14

    Context layer for AI agents that compiles enterprise knowledge into a governed, verified source of truth and scores agent responses against it, with an AI visibility benchmarking product.

    Full coverage on Observability & Auditability and Testing, Debugging & Optimization and 3 more. No documented coverage on Browser / Computer-use and Deployment & Data Residency.

65. 65

    Straiker8.5 / 14

    Agentic-first AI security platform securing agents across their lifecycle through three modules: Discover AI (inventory of agents, tools, and MCP servers with shadow-usage detection), Ascend AI (continuous adversarial testing on every deployment, mapped to OWASP LLM Top 10, MITRE ATLAS, and the EU AI Act), and Defend AI (sub-second runtime blocking of prompt injection, agent manipulation, and regulated-data leakage). Single-hook install, ISO 27001 and SOC 2 certified, ~$85M raised, customers include Comcast, Fortinet, Snowflake, and Deloitte.

66. 66

    WitnessAI8.5 / 14

    Unified AI security and governance platform applying intent-based behavioral controls at the network layer: analyzes the meaning behind each prompt to block multi-turn attacks and contextual jailbreaks, monitors which agents are active and what MCP servers, tools, and data they touch, and connects human and agentic identities for explainable agent actions across both workforces. Deploys on-premise, in a cloud sandbox, or in a VPC. Over $85M raised (Sound Ventures, GV, Qualcomm, Samsung), 500% ARR growth, and production customers among the largest public enterprises.

67. 67

    Anchor Browser8.0 / 14

    Cloud hosted browser infrastructure that lets AI agents operate real web interfaces, navigating, authenticating, filling forms, and extracting data on sites that have no API.

    Full coverage on Security, Identity & Governance and Integrations & Tool Calling and 3 more. No documented coverage on Human Oversight & Guardrails and Testing, Debugging & Optimization and 1 more.

68. 68

    Arcade8.0 / 14

    Authenticated tool calling platform and MCP runtime that handles OAuth and token management so AI agents can securely act across 7,000+ integrations.

    Full coverage on Integrations & Tool Calling and Security, Identity & Governance and 3 more. No documented coverage on Knowledge Grounding & RAG and Memory & State Persistence and 1 more.

69. 69

    Beam8.0 / 14

    Open source serverless sandboxes and GPU cloud for production agents, with gVisor isolation, stateful snapshots, and bring your own compute.

70. 70

    Braintrust8.0 / 14

    AI evaluation and observability platform with self-serve pricing, MCP support, human eval loops, dataset management, and Series B traction. Infrastructure for eval-driven AI development teams.

71. 71

    Causa Prima8.0 / 14

    Agent to agent network for B2B finance where buyer and supplier AI agents settle invoices, disputes, and payment terms directly in real time, aiming to retire fixed net payment terms.

72. 72

    Composio8.0 / 14

    Agent integrations layer providing 1,000+ app connections with delegated auth, MCP positioning, reusable toolkits, and SOC 2 / ISO 27001 compliance. Infrastructure for agent tool access, not an end-user assistant.

73. 73

    Dome Systems8.0 / 14

    Agent native governance control plane that registers agents, filters tools via an MCP Gateway, enforces per action Cedar policy before execution, and audits every governed action across any cloud.

    Full coverage on APIs / SDKs / MCP Extensibility and Human Oversight & Guardrails and 4 more. No documented coverage on Testing, Debugging & Optimization and Browser / Computer-use and 2 more.

74. 74

    GPT Researcher8.0 / 14

    Open-source autonomous research agent with planner/execution multi-agent architecture and cited reports.

    Full coverage on Workflow Orchestration and Knowledge Grounding & RAG and 3 more. No documented coverage on Human Oversight & Guardrails and Memory & State Persistence and 1 more.

75. 75

    Kernel8.0 / 14

    Browsers as a service for AI agents: millisecond spin ups, persistent session state, live take-control views and replays, any-framework support plus an MCP server, and an agent runtime that turns agents into HTTPS endpoints. $22M from Accel and YC.

76. 76

    Modal8.0 / 14

    Python native serverless cloud for AI, with sub second cold starts, per second GPU billing, and isolated sandboxes for running untrusted agent code at scale.

77. 77

    Monte Carlo8.0 / 14

    Agent trust platform that unifies data and agent observability, so enterprise teams can monitor, troubleshoot, and improve production AI systems from the pipelines feeding agents to the outputs they produce.

78. 78

    Portkey8.0 / 14

    LLM Gateway and governance platform with MCP Gateway support, audit logs, routing, and enterprise security/compliance. Infrastructure for platform teams managing LLM and agent traffic.

79. 79

    QA Wolf8.0 / 14

    Agentic automated testing platform and managed service, Coverage as a Service, where AI agents map an app and generate Playwright and Appium tests while full time human QA engineers build, run, and maintain the suite with a Zero Flake Guarantee, delivering eighty percent plus E2E coverage integrated into CI.

80. 80

    Steel8.0 / 14

    Open source browser API for AI agents that runs cloud or self hosted Chromium sessions with anti detection, session persistence, and a live session viewer.

81. 81

    Unikraft8.0 / 14

    Millisecond compute layer that runs AI agent sandboxes and headless browsers as strongly isolated microVMs, with instant boot, checkpoint and fork, and scale to zero.

    Full coverage on Deployment & Data Residency and Memory & State Persistence and 3 more. No documented coverage on Human Oversight & Guardrails and Knowledge Grounding & RAG and 1 more.

82. 82

    Zenity8.0 / 14

    First end-to-end security and governance platform purpose-built for AI agents: Observe (discovery and observability), Govern (AI Security Posture Management), and Defend (detection and response) across Copilot Studio, Agentforce, Bedrock, Azure AI Foundry, Vertex AI, and ChatGPT Enterprise, with intent-aware runtime enforcement. Named Gartner's company to beat in AI agent governance.

83. 83

    Arize AI7.5 / 14

    AI observability and evaluation platform that traces, evaluates, and monitors LLM applications and agents, with an open source core in Phoenix.

    Full coverage on APIs, SDKs & MCP Extensibility and Testing, Debugging & Optimization and 3 more. No documented coverage on Browser & Computer Use and Workflow Orchestration and 2 more.

84. 84

    CalypsoAI7.5 / 14

    Enterprise AI security platform that red teams, defends, and monitors AI models, apps, and agents at the inference layer. Now an F5 company.

    Full coverage on Testing, Debugging & Optimization and Human Oversight & Guardrails and 3 more. No documented coverage on Browser & Computer Use and Workflow Orchestration and 2 more.

85. 85

    Capsule Security7.5 / 14

    Runtime security layer for AI agents that detects and blocks unsafe agent behavior in real time, with agentless discovery, an agent security graph, and white box agent red teaming.

86. 86

    Daytona7.5 / 14

    Open source infrastructure that spins up secure, isolated sandboxes in under 90 milliseconds so AI agents can run generated code safely.

    Full coverage on Memory & State Persistence and Deployment & Data Residency and 2 more. No documented coverage on Knowledge Grounding & RAG and Observability & Auditability and 1 more.

87. 87

    Evoke Security7.5 / 14

    Security platform for the agentic workforce, an EDR for AI agents that discovers every agent, model, tool and MCP, models attack paths, and blocks risky agent actions in real time.

    Full coverage on Deployment & Data Residency and Human Oversight & Guardrails and 3 more. No documented coverage on Browser / Computer-use and Model Flexibility & Routing and 2 more.

88. 88

    Fiddler AI7.5 / 14

    Enterprise AI observability and security control plane for agents, LLM apps, and ML models, with real time guardrails, explainability, and governance.

89. 89

    Hyperspell7.5 / 14

    API based memory and context layer that connects company tools into a live context graph any AI agent can read from and write back to, with every fact traced to source.

90. 90

    Jetty7.5 / 14

    Infrastructure for running agentic AI workflows: one OpenAI-compatible endpoint that proxies over 100 LLM providers, runs autonomous agents in isolated sandboxes from markdown runbooks, and applies evaluation loops with full trajectory tracing.

    Full coverage on Workflow Orchestration and Human Oversight & Guardrails and 4 more. No documented coverage on Deployment & Data Residency and Knowledge Grounding & RAG and 3 more.

91. 91

    Langbase7.5 / 14

    Serverless platform to build, deploy, and scale composable AI agents with built in memory, tools, and a unified API to hundreds of LLMs.

    Full coverage on Integrations & Tool Calling and Knowledge Grounding & RAG and 2 more. No documented coverage on Human Oversight & Guardrails and Triggers & Channel Coverage and 1 more.

92. 92

    Langfuse7.5 / 14

    Open-source AI engineering platform for tracing, observability, prompt management, and evaluation, self-hostable or SaaS, with an OSS community, free plan, and model-agnostic coverage.

93. 93

    Personal AI7.5 / 14

    Memory infrastructure for AI that uses an identity-based, persistent memory architecture and custom small language models to let enterprises train and deploy their own AI personas, or teammates, grounded in proprietary knowledge.

94. 94

    Toolhouse7.5 / 14

    Platform to build, deploy, and run AI agents from a prompt, with built in RAG, tools, and one link MCP connectivity to thousands of services.

    Full coverage on APIs, SDKs & MCP Extensibility and Integrations & Tool Calling and 1 more. No documented coverage on Browser & Computer Use and Human Oversight & Guardrails.

95. 95

    AIsa7.0 / 14

    Unified resource and payment gateway for AI agents that lets them discover, access, and pay for models, APIs, data, and compute through one programmable interface, with usage based billing settled in fiat or stablecoins.

96. 96

    Confident AI7.0 / 14

    LLM evaluation and observability platform from the creators of DeepEval, with 50+ open source metrics for testing agents, RAG, and chatbots.

    Full coverage on Observability & Auditability and Deployment & Data Residency and 2 more. No documented coverage on Workflow Orchestration and Knowledge Grounding & RAG and 2 more.

97. 97

    E2B7.0 / 14

    Secure code sandbox infrastructure for AI agents: isolated virtual-computer execution environments, SDKs, and PAYG pricing for AI developers who need safe, ephemeral code execution. Infrastructure component, not an end-user agent.

98. 98

    Hyperbrowser7.0 / 14

    Cloud headless browser infrastructure for AI agents with stealth, proxy rotation, and CAPTCHA solving, plus the open source HyperAgent framework and an MCP server.

99. 99

    LiteLLM7.0 / 14

    Open source AI gateway and Python SDK giving one OpenAI compatible interface to more than a hundred model providers, with keys, budgets, routing, and self hosting.

100. 100

     Opik7.0 / 14

     Open source LLM and agent observability and evaluation platform from Comet, with tracing, LLM as a judge metrics, guardrails, and an agent optimizer.

101. 101

     Trent AI7.0 / 14

     Multi agent security platform that secures a company's own AI agents across their lifecycle, using specialist scan, judge, mitigate, and evaluate agents to catch and fix threats like prompt injection, tool misuse, and data exfiltration.

102. 102

     W&B Weave7.0 / 14

     Agent and LLM observability, evaluation, and guardrails from Weights and Biases, with session aware tracing, LLM as a judge scoring, and safety scorers.

103. 103

     Apaleo6.5 / 14

     Open, API first hospitality property management platform pivoting into agentic infrastructure, with the Agent Hub marketplace of hireable AI agents, the first hospitality MCP Server built on Anthropic's Model Context Protocol so agents can run cross system hotel operations without custom integrations, and an Agent Factory for co creating agents.

104. 104

     BAND6.5 / 14

     Universal communication and coordination layer for AI agents, a service mesh for agents, that lets agents built in any framework discover, delegate, and collaborate with each other and with humans in shared rooms, with built in governance that verifies authority and tracks who did what under which policy.

105. 105

     Bright Data Agent Browser6.5 / 14

     Bright Data's cloud browser for agents: GUI Chromium fleets on 400M+ rotating IPs with autonomous CAPTCHA solving across every major challenge system, drop-in CDP compatibility, an MCP server, per-session fingerprints, and audit logs. The unlocking layer as agent infrastructure.

106. 106

     Clutch Security6.5 / 14

     Universal non human identity security platform that discovers, governs, and secures every NHI, AI agent, and secret across cloud, SaaS, and on prem, powered by the Identity Lineage graph and a zero trust, ephemeral credentials approach.

107. 107

     Cognee6.5 / 14

     Open source AI memory platform giving agents persistent long term memory through a hybrid graph, vector, and relational engine with a memory native API.

108. 108

     Galileo6.5 / 14

     AI reliability platform that evaluates, observes, and guardrails GenAI apps and agents using its own low latency Luna evaluation models.

     Full coverage on Human Oversight & Guardrails and Observability & Auditability and 2 more. No documented coverage on Workflow Orchestration and Knowledge Grounding & RAG and 3 more.

109. 109

     Gentoro6.5 / 14

     MCP platform that uses LLMs to turn OpenAPI specs into secure, agent ready MCP tools and govern them across their lifecycle.

     Full coverage on Integrations & Tool Calling and Security, Identity & Governance and 2 more. No documented coverage on Knowledge Grounding & RAG and Memory & State Persistence and 3 more.

110. 110

     HoneyHive6.5 / 14

     OpenTelemetry native platform to trace, evaluate, and monitor AI agents across development and production, combining automated and human evaluation with a versioned system of record.

     Full coverage on Security, Identity & Governance and Observability & Auditability and 2 more. No documented coverage on Workflow Orchestration and Knowledge Grounding & RAG and 3 more.

111. 111

     Kerno6.5 / 14

     Self building backend test harness that validates every code change including AI generated changes against the real stack in the IDE, self healing the suite as the codebase evolves.

     Full coverage on Deployment & Data Residency and Testing, Debugging & Optimization. No documented coverage on Model Flexibility & Routing and Prebuilt Agents / Templates / Packs and 1 more.

112. 112

     LangWatch6.5 / 14

     Open source platform for agent observability, evaluation, and scenario based simulation, with OpenTelemetry tracing and a built in AI gateway.

113. 113

     Maxim AI6.5 / 14

     End to end platform to simulate, evaluate, and observe AI agents across the lifecycle, with a built in gateway to a thousand plus models.

     Full coverage on Observability & Auditability and Model Flexibility & Routing and 1 more. No documented coverage on Workflow Orchestration and Knowledge Grounding & RAG and 2 more.

114. 114

     OpenBox AI6.5 / 14

     Runtime governance and trust layer for enterprise AI agents that enforces identity, authorization, and policy at the point of execution, scores risk against the OWASP AIVSS framework, and produces tamper proof cryptographic Proof Certificates, dropping into Temporal, LangChain, Mastra, and other runtimes as a plugin.

115. 115

     Prime Intellect6.5 / 14

     Open infrastructure for training and running AI agents, pairing a global GPU marketplace and inference API with Lab, an end to end platform for reinforcement learning, evaluation, and deployment across more than two thousand five hundred open environments.

116. 116

     Rye6.5 / 14

     Agentic commerce infrastructure whose Universal Checkout API turns a product URL and tokenized payment into a confirmed order on any merchant through AI browser automation, covering fifteen thousand plus stores with high reliability, enforced spend controls, and a companion Returns and Product Data API.

117. 117

     SGNL6.5 / 14

     Continuous Identity platform delivering dynamic, context aware authorization and zero standing privilege for human, non human, and AI agent identities, with a built in MCP Gateway governing what tools agents can use; being acquired by CrowdStrike in a deal announced January 2026.

118. 118

     Temporal Cortex6.5 / 14

     Open scheduling infrastructure that gives any AI agent reliable calendar scheduling over MCP, A2A, and REST, with atomic booking and deterministic time logic.

     Full coverage on Integrations & Tool Calling and Observability & Auditability and 2 more. No documented coverage on Browser / Computer-use and Knowledge Grounding & RAG and 3 more.

119. 119

     AgentOps6.0 / 14

     Agent observability and reliability platform with broad model and framework support, OSS SDK surface, and devtools for production-agent debugging. An infrastructure component, not an end-user assistant.

120. 120

     Cyara6.0 / 14

     Established CX assurance platform that added agentic testing and AI governance so enterprises can validate, monitor, and control customer facing AI agents, using AI test agents to test other agents across voice and digital channels and AI Trust modules to catch compliance, bias, and hallucination risks before customers do.

121. 121

     Deepgram6.0 / 14

     Voice AI infrastructure with real time speech to text (Nova, Flux), text to speech (Aura), and a Voice Agent API that unifies listening, thinking, and speaking.

122. 122

     Firecrawl6.0 / 14

     Open source web scraping and crawling API that turns any site into clean, LLM ready markdown or structured JSON for agents.

     Full coverage on Knowledge Grounding & RAG and Deployment & Data Residency and 1 more. No documented coverage on Human Oversight & Guardrails and Observability & Auditability and 3 more.

123. 123

     Inth6.0 / 14

     Code native privacy and consent governance built on the open source c15t SDK, extending to govern how AI systems and agents handle user data.

     Full coverage on Observability & Auditability and APIs / SDKs / MCP Extensibility and 1 more. No documented coverage on Model Flexibility & Routing and Knowledge Grounding & RAG and 3 more.

124. 124

     Klavis AI6.0 / 14

     MCP infrastructure with hosted, production MCP servers and built in OAuth across 600 plus tools, plus the open source Strata server and agent training environments.

125. 125

     Patronus AI6.0 / 14

     AI evaluation, guardrails, and agent debugging platform with research grade models for hallucination detection, embedded through an API to catch failures before users do.

     Full coverage on Human Oversight & Guardrails and Testing, Debugging & Optimization. No documented coverage on Workflow Orchestration and Knowledge Grounding & RAG and 2 more.

126. 126

     Prefactor6.0 / 14

     Framework agnostic control plane for governing an AI agent portfolio with identity, quality gates, cost attribution, and lifecycle governance.

     Full coverage on Security, Identity & Governance and Human Oversight & Guardrails and 2 more. No documented coverage on Prebuilt Agents / Templates / Packs and Model Flexibility & Routing and 4 more.

127. 127

     SageOx6.0 / 14

     Shared context infrastructure for AI-native teams that captures decisions, intent, and history from conversations, chats, and coding sessions and feeds it as persistent, structured memory into every human and AI agent so they stay aligned.

128. 128

     TectoAI6.0 / 14

     Governance platform for agentic AI: discover, vet, and continuously monitor third party AI tools and agents, with compliance controls for regulated industries.

129. 129

     Cartesia5.5 / 14

     Real time voice models (Sonic text to speech, Ink speech to text) and the Line voice agent platform, built on state space models for sub 100 millisecond latency.

130. 130

     Catalog5.5 / 14

     Product data layer for AI commerce that structures and enriches a brand's catalog into machine readable data and syndicates it, via MCP, to the AI surfaces where shoppers ask.

     Full coverage on Knowledge Grounding & RAG and Integrations & Tool Calling and 2 more. No documented coverage on Security, Identity & Governance and Workflow Orchestration and 5 more.

131. 131

     Freeplay5.5 / 14

     LLM product testing, evaluation, and observability platform that unifies prompt management, evals, experiments, and production monitoring for cross functional teams.

     Full coverage on Testing, Debugging & Optimization and Observability & Auditability. No documented coverage on Browser & Computer Use and Workflow Orchestration and 3 more.

132. 132

     Guardrails AI5.5 / 14

     Open source framework for validating and correcting LLM outputs with composable guardrails plus a hub of prebuilt validators for safety and structure.

     Full coverage on Human Oversight & Guardrails and Deployment & Data Residency. No documented coverage on Workflow Orchestration and Knowledge Grounding & RAG and 3 more.

133. 133

     Hamming AI5.5 / 14

     Voice and chat agent QA platform that simulates thousands of test calls and monitors production calls with 50 plus voice specific metrics.

     Full coverage on Observability & Auditability and Testing, Debugging & Optimization. No documented coverage on Workflow Orchestration and Knowledge Grounding & RAG and 3 more.

134. 134

     Helicone5.5 / 14

     Open source LLM observability platform and AI gateway that logs, routes, caches, and costs every request through one OpenAI compatible endpoint.

     Full coverage on Model Flexibility & Routing and Observability & Auditability and 1 more. No documented coverage on Browser & Computer Use and Workflow Orchestration and 4 more.

135. 135

     Lemrock5.5 / 14

     Paris deeptech building vendor neutral middleware for agentic commerce: a single integration ingests a brand's catalog, structures it for large language models, and makes products discoverable and transactable inside AI assistants like ChatGPT, Claude, Gemini, and Perplexity while brands keep control of pricing and margins.

136. 136

     Mem05.5 / 14

     Open source memory layer that gives AI agents persistent, personalized recall across sessions through a simple add and search API, cutting token costs.

     Full coverage on Memory & State Persistence and Deployment & Data Residency and 1 more. No documented coverage on Workflow Orchestration and Human Oversight & Guardrails and 4 more.

137. 137

     Neurelo5.5 / 14

     Cloud platform whose Neurelo Connect gives AI agents secure, deterministic, real-time access to PostgreSQL, MongoDB, MySQL, and ClickHouse via auto-generated REST and GraphQL APIs and MCP.

     Full coverage on Integrations & Tool Calling and Security, Identity & Governance and 1 more. No documented coverage on Prebuilt Agents / Templates / Packs and Browser / Computer-use and 4 more.

138. 138

     Rime5.5 / 14

     Rime is enterprise text to speech infrastructure, not a full agent, providing conversational voice models that power high volume phone agents for brands like Domino's, with on premises deployment.

     Full coverage on Security, Identity & Governance and Deployment & Data Residency and 2 more. No documented coverage on Knowledge Grounding & RAG and Human Oversight & Guardrails and 5 more.

139. 139

     Smithery5.5 / 14

     Registry and hosting platform for thousands of MCP servers, with local CLI install, one click hosting, and Toolbox, a meta MCP that routes agents to the right tools.

140. 140

     Vocode5.5 / 14

     Vocode is an open source developer framework for building voice agents on any AI stack, offering Python and Node libraries, a hosted calling service, and full model flexibility.

     Full coverage on Deployment & Data Residency and Model Flexibility & Routing and 1 more. No documented coverage on Knowledge Grounding & RAG and Human Oversight & Guardrails and 4 more.

141. 141

     Zep5.5 / 14

     Memory and context engineering platform for AI agents built on a temporal knowledge graph that tracks how facts change over time.

     Full coverage on Memory & State Persistence and Deployment & Data Residency and 1 more. No documented coverage on Workflow Orchestration and Human Oversight & Guardrails and 4 more.

142. 142

     Exa5.0 / 14

     Neural web search API built for AI agents and RAG that retrieves pages by meaning and returns token efficient, LLM ready contents.

     Full coverage on APIs, SDKs & MCP Extensibility and Knowledge Grounding & RAG. No documented coverage on Model Flexibility & Routing and Testing, Debugging & Optimization and 4 more.

143. 143

     Lakera5.0 / 14

     AI native security platform that protects LLMs and agents from prompt injection, jailbreaks, and data leakage at runtime, with continuous red teaming.

     Full coverage on Human Oversight & Guardrails and Security, Identity & Governance. No documented coverage on Workflow Orchestration and Knowledge Grounding & RAG and 4 more.

144. 144

     OpenRouter5.0 / 14

     Unified API and marketplace routing agents to 400 plus models through one OpenAI compatible endpoint, with failover, cost controls, and pass through pricing.

145. 145

     AgentQL4.5 / 14

     A self-healing query language for the web: natural language selectors that find elements by meaning, survive layout changes, and return structured JSON or Playwright locators. SDKs, REST API, MCP server, and framework integrations — the semantic selector layer beneath agents.

146. 146

     Circuit & Chisel4.0 / 14

     Agentic payments infrastructure company building ATXP, an open, web-wide protocol that lets AI agents discover, negotiate, and settle instant low-cost micropayments with each other, connecting agentic commerce, stablecoins, and payments into one rail.

147. 147

     Jina AI4.0 / 14

     Search foundation platform with a Reader, embeddings, reranker, and DeepSearch that grounds agents and RAG systems in clean, retrievable web data.

148. 148

     Tavily4.0 / 14

     Real time web search, extraction, crawling, and research API built for AI agents, returning clean LLM ready content that keeps agents grounded.

     Full coverage on Integrations & Tool Calling and Knowledge Grounding & RAG and 1 more. No documented coverage on Human Oversight & Guardrails and Security, Identity & Governance and 7 more.

149. 149

     Voker3.5 / 14

     Agent analytics platform, the Amplitude for agents, whose provider agnostic SDK turns user to agent conversations into structured analytics, automatically detecting intents, corrections, and resolutions and tying agent behavior to business outcomes so whole product teams can measure and improve agents without reading traces.

150. 150

     Lightpanda3.0 / 14

     A headless browser engine written from scratch in Zig for agents: ~11x faster and ~9x lighter than headless Chrome, CDP compatible with Playwright and Puppeteer, open source with a drop-in managed cloud. The engine layer of the agent browser stack; currently beta.

151. 151

     Acrab2.0 / 14

     Singapore compute infrastructure company building a full stack platform for agentic AI at the edge, spanning its own AI silicon, local LLM inference, an operating system, multimodal interfaces, and agent orchestration, with a first generation platform called GELIX moving toward mass production.