AI visibility report
Arcade.dev ranks #2 in Agent Authentication & Identity for AI AI search.
Outside the top three on 10 of the 25 prompts buyers actually ask.
WorkOS is cited on 9 of those losses.
Free trial. Setup comes pre-filled for Arcade.dev.
Track Arcade.dev across these prompts daily.
Start free trial#2 among 13 vendors · still absent from 75.3% of tracked prompt responses
Top-3 citations across 150 prompt × platform pairs
Peer Ranking
Key Metrics
Platform Breakdown
Perplexity
Bing Copilot
Gemini Search
Grok
ChatGPT
Google AI ModeVisible, but narrative can improve. Arcade.dev ranks #2 on presence but #6 on sentiment. The brand appears relatively often, but competitors may be getting more favorable language when they appear.
Where Arcade.dev is losing
Prompts where competitors are visible and Arcade.dev is not.
These prompt-level losses are the first prompts to track and repair.
Where Arcade.dev is winning3
Which platforms offer the smoothest workflow for testing and debugging agent auth flows during development?
Avg # 1.0 · 1 platform
I'm evaluating agent auth tools for a multi-cloud setup — which ones support cross-environment policy enforcement and audit trails?
Avg # 1.3 · 3 platforms
Looking for agent identity infrastructure that plugs into our existing identity provider — which platforms support federation?
Avg # 7.0 · 1 platform
Where Arcade.dev is losing5
What's the quickest way to add OAuth authentication to AI agents that need to access third-party APIs on behalf of users?
Competitors on 5 platforms
Track this promptWhich agent auth platforms have the widest coverage of pre-built OAuth connectors for popular SaaS APIs?
Competitors on 4 platforms
Track this promptWhich platforms make it easiest to set up secure auth for MCP servers without building custom OAuth flows from scratch?
Competitors on 4 platforms
Track this promptLooking for a drop-in SDK to handle agent-to-API authentication in a TypeScript codebase — what are my options?
Competitors on 3 platforms
Track this promptWhat tools integrate natively with MCP and the major AI agent frameworks for handling auth in agentic workflows?
Competitors on 3 platforms
Track this prompt
Track Arcade.dev daily before the next report refresh.
Track these gapsResearch dossierCapabilities, use cases, sources, reviews, pricing, and FAQ
Overview
Arcade.dev is a San Francisco-based AI infrastructure company that provides an MCP (Model Context Protocol) runtime for deploying production-grade AI agents. Founded in February 2024 by Alex Salazar (former Okta VP and Stormpath co-founder) and Sam Partee (former Principal AI Engineer at Redis), Arcade solves the core bottleneck in enterprise AI agent adoption: secure, per-user authentication and reliable tool integration at scale. Its platform supplies agent-optimized tools across 40+ major SaaS systems, manages OAuth token lifecycles without exposing credentials to LLMs, and offers centralized governance via MCP Gateways. Arcade is compatible with all major MCP clients and LLM frameworks, deploys in cloud or on-prem environments, and holds SOC 2 certification. The company raised a $12M seed round in March 2025.
Arcade.dev is an MCP runtime platform that enables AI agents to securely take real-world actions on behalf of users across enterprise and consumer applications. It provides authenticated tool-calling infrastructure—handling per-user OAuth flows, token management, and credential security—so that agents built on any LLM framework can interact with systems like Gmail, Slack, Salesforce, GitHub, and 40+ others without service-account risk or bespoke integration work. The platform includes a curated catalog of agent-optimized MCP tools, an open-source SDK for building custom tools, MCP Gateways for team-level governance, and enterprise controls (RBAC, audit logs, flexible deployment) that satisfy security and compliance requirements.
Key Facts
- Founded
- 2024
- HQ
- San Francisco, CA, USA
- Founders
- Alex Salazar, Sam Partee
- Employees
- 20-50
- Funding
- $12M
- Status
- Private
Target users
Key Capabilities10
- MCP runtime for multi-user, production-grade AI agent deployments
- Per-user OAuth authorization (agents act as the user, not a shared service account)
- Secure token management—LLMs never access raw credentials
- Pre-built catalog of 8,000+ agent-optimized MCP tool integrations
- Open-source Secure MCP Framework SDK (Python/JavaScript) for custom tool creation
- MCP Gateway: single-URL, curated tool collections with centralized governance and RBAC
- Agent lifecycle governance: audit logs, observability, compliance reporting
- Flexible deployment: cloud-hosted, VPC, on-premises, or fully air-gapped
- Bring-Your-Own-Credentials (BYOC) support for Pro tool executions
- SOC 2 certified; SSO/SAML and RBAC for enterprise plans
Key Use Cases7
- Deploying production multi-user AI agents that take authenticated actions across SaaS apps
- AI sales agents reading CRM data and sending emails or updating records on behalf of reps
- AI assistants in Slack or Teams that access Gmail, Calendar, GitHub, and other tools
- Agentic e-commerce workflows: search, compare, and checkout with AI-managed payments
- Enterprise platform teams governing MCP tool access across engineering, sales, and support personas
- AI coding agents in Cursor or VS Code with authenticated access to GitHub and Linear
- Internal developer tooling to build and deploy custom MCP servers with built-in OAuth
Arcade.dev customer outcomes
Using Arcade, YUR authenticated users' Twitter and LinkedIn accounts for their AI agent without managing refresh tokens, broken auth, or OAuth setup complexity, eliminating a major integration burden.
Sybill built an AI sales agent that analyzes calls, emails, and CRM data; with Arcade the agent takes secure actions in those systems on behalf of sales reps, turning it from an assistant into a revenue-driving tool.
Eddo Learning bypassed Google service integration complexity using Arcade, enabling faster development of AI assistant tools for teachers.
Recent Trend
How AI describes Arcade.dev3
arcade.dev * Delegation Tokens / Verifiable Credentials: When a user initializes an agent, they issue a cryptographically signed token that strictly scopes what the agent can do (e.g., read a specific folder, but never transfer funds).
What auth infrastructure holds up when thousands of AI agents are making concurrent authenticated API calls?
arcade.dev ### 2\. Standard Identity & Vault Solutions For teams building custom architectures or handling enterprise-grade governance, standard auth platforms and secret management layers are adapted for agent use: * Scalekit / WorkOS / Styt...
What tools do AI agent developers actually use day-to-day for handling user-delegated authentication across multiple integrations?
arcade.dev * Multi-Hop Context Propagation: The capability to track and limit how deep a delegation chain can go (preventing endless loops or privilege escalation).
Looking for an agent auth solution that supports agent-to-agent delegation and MCP server authentication — what should I evaluate?
Most cited sources8
1327 Best AI Agent Authentication Platforms (2026) | Arcade
arcade.dev·Blog Post
- 21
Multi-User AI Agent Auth: OAuth & MCP Guide | Arcade.dev
arcade.dev·Blog Post
- 2
SSO for AI Agents: Authentication and Authorization Guide
arcade.dev·Blog Post
- 1
AI Agent & MCP Development Blog: Make AI Do Work
arcade.dev·Blog Post
- 1
MCP Toolkit Benchmark: Arcade.dev vs Composio
arcade.dev·Blog Post
- 1
AgentKit + MCP: Solving the Auth Gap for Production Agents
arcade.dev·Blog Post
Alternatives in Agent Authentication & Identity for AI6
Arcade.dev positions itself as the only purpose-built MCP (Model Context Protocol) runtime for production AI agents, differentiating on three axes: (1) authentication-first design using per-user OAuth rather than shared service accounts, preventing LLMs from ever seeing raw tokens; (2) a pre-built catalog of 8,000+ agent-optimized tool integrations that are more than raw API wrappers; and (3) enterprise deployment flexibility (cloud, VPC, on-prem, air-gapped).
- Its founding team alumni from Okta/Stormpath and Redis gives it unique credibility at the intersection of auth and AI infra—a combination that general auth vendors (Auth0, Stytch, WorkOS) and integration-layer competitors (Composio, Nango) do not replicate.
- Arcade targets the gap where MCP adoption stalls: moving from localhost demos to secure, multi-user, enterprise-grade agent deployments.
Reviews
Praised
- Eliminates OAuth/refresh-token complexity for agent developers
- Single-command path from local prototype to production MCP server
- Strong developer experience and clean Python SDK
- Agents act as the end user, not a shared service account
- Large and growing catalog of agent-optimized integrations
- Easy composability across LLM frameworks and architectures
- SOC 2 certification and enterprise security posture
Criticized
- Pro tool execution costs can compound at scale
- Enterprise features (RBAC, audit logs, SSO) locked to custom-priced tier
- Broader MCP ecosystem still maturing, creating protocol-level friction
- Limited third-party review coverage—early-stage brand recognition
- 99% of existing MCP servers are single-user, limiting interoperability
Arcade.dev lacks a substantial presence on mainstream review platforms (G2, Gartner Peer Insights) as of early 2025, consistent with its early-stage status. Developer sentiment gathered from Product Hunt, technical blog commentary, and named customer testimonials is strongly positive. Practitioners highlight the elimination of OAuth refresh-token headaches, the single-command path from local prototype to production MCP server, and the quality of framework integrations. LangChain's founding engineer described it as 'the best platform to facilitate secure and interactive MCP.' Snyk's Head of Developer & Security Relations called it 'insanely useful—finally a product that lets AI agents actually do stuff.' The most cited friction points in early developer commentary relate to the broader MCP ecosystem immaturity and the cost structure of Pro tool executions at scale.
Pricing
Usage-based pricing across three tiers. Hobby (free): 100 user authorization challenges, 1,000 standard tool executions, 50 pro tool executions, 1 Arcade-hosted MCP server included. Growth ($25/month): 600 user challenges (then $0.05 each), 2,000 standard executions (then $0.01 each), 100 pro executions (then $0.50 each), hosted MCP servers at $0.05/hour, email SLA support. Enterprise (custom): dedicated tenant isolation, audit logs, RBAC, SSO/SAML, dedicated account rep, custom SLAs. A startup program with discounted access is available for companies under 100 employees, nonprofits, and educational institutions. Bring-Your-Own-Credentials (BYOC) reduces pro tool execution costs.
Limitations
- MCP protocol itself still maturing—Arcade's docs note that 99% of existing MCP servers are single-user and lack robust multi-user auth; Arcade's runtime bridges this gap but enterprises face broader ecosystem immaturity.
- Pro tool executions carry additional per-call costs that can compound at scale.
- Dedicated tenant isolation and compliance features (audit logs, RBAC, SSO/SAML) are Enterprise-tier only.
- Bring-Your-Own-Credentials is required for some Pro tools to control costs.
- Early-stage company (founded February 2024) with a relatively small team, which may present enterprise support and SLA concerns for the largest deployments.
- No verified presence on G2 or Gartner Peer Insights as of Q2 2025, limiting third-party review signal.
Frequently asked questions
Topic coverageCoverage by buyer topic
Topic Coverage
Prompt-Level Results
| Prompt | Bing Copilot | Perplexity | Gemini Search | Google AI Mode | ChatGPT | Grok |
|---|---|---|---|---|---|---|
Capability2/5 cited (40%) | ||||||
Looking for an agent auth solution that supports agent-to-agent delegation and MCP server authentication — what should I evaluate? | Neither your brand nor a competitor was cited | A competitor was cited | Your brand was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Your brand and a competitor were cited |
Which platforms support ephemeral, task-scoped credentials for AI agents instead of static API keys? | Neither your brand nor a competitor was cited | A competitor was cited | A competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
What tools handle both human user auth and AI agent identity under a single platform with granular per-agent permissions? | Your brand was cited | Your brand and a competitor were cited | A competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited |
Which non-human identity platforms can govern AI agents alongside service accounts, API keys, and machine identities? | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
What tools let me enforce fine-grained, intent-based access policies for AI agents accessing enterprise systems? | Neither your brand nor a competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
Developer Experience5/5 cited (100%) | ||||||
What tools do AI agent developers actually use day-to-day for handling user-delegated authentication across multiple integrations? | Your brand and a competitor were cited | Your brand and a competitor were cited | Your brand and a competitor were cited | Neither your brand nor a competitor was cited | A competitor was cited | Your brand and a competitor were cited |
Which platforms offer the smoothest workflow for testing and debugging agent auth flows during development? | Neither your brand nor a competitor was cited | Your brand and a competitor were cited | A competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
Which agent auth platforms have the best developer experience for managing OAuth tokens, refresh flows, and scoped permissions? | A competitor was cited | Your brand and a competitor were cited | Your brand and a competitor were cited | A competitor was cited | A competitor was cited | Your brand and a competitor were cited |
I'm an ML engineer building agents — which auth platforms let me focus on agent logic instead of wrestling with OAuth plumbing? | Your brand was cited | Your brand and a competitor were cited | Neither your brand nor a competitor was cited | A competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited |
What agent identity tools have the best docs and SDKs for a small team building their first production AI agent? | Neither your brand nor a competitor was cited | Your brand and a competitor were cited | Neither your brand nor a competitor was cited | A competitor was cited | A competitor was cited | A competitor was cited |
Integrations & Ecosystem4/5 cited (80%) | ||||||
Looking for agent identity infrastructure that plugs into our existing identity provider — which platforms support federation? | Neither your brand nor a competitor was cited | Your brand was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
Which agent auth platforms have the widest coverage of pre-built OAuth connectors for popular SaaS APIs? | Your brand and a competitor were cited | Your brand and a competitor were cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited | Your brand and a competitor were cited |
What tools integrate natively with MCP and the major AI agent frameworks for handling auth in agentic workflows? | Neither your brand nor a competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited | A competitor was cited |
I'm evaluating agent auth tools for a multi-cloud setup — which ones support cross-environment policy enforcement and audit trails? | Your brand was cited | Your brand and a competitor were cited | Your brand was cited | A competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited |
Which platforms let AI agents securely access both cloud SaaS tools and on-prem internal systems through one auth layer? | Neither your brand nor a competitor was cited | Your brand and a competitor were cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
Performance & Reliability4/5 cited (80%) | ||||||
What auth infrastructure holds up when thousands of AI agents are making concurrent authenticated API calls? | Your brand was cited | Your brand and a competitor were cited | Your brand and a competitor were cited | A competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited |
Which agent auth platforms add the least latency overhead to tool calls when agents need to authenticate in real time? | Your brand was cited | Your brand and a competitor were cited | A competitor was cited | A competitor was cited | Your brand was cited | Your brand and a competitor were cited |
Which platforms handle automatic token refresh and rotation reliably enough for production AI agent workloads? | Neither your brand nor a competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | Your brand and a competitor were cited | A competitor was cited | Neither your brand nor a competitor was cited |
What are the most battle-tested platforms for securing AI agents in regulated industries like fintech or healthcare? | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
I need agent auth that works at the edge with sub-100ms enforcement — which platforms support distributed authorization? | Neither your brand nor a competitor was cited | Your brand and a competitor were cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
Setup & First Run2/5 cited (40%) | ||||||
I'm building an AI agent that needs to connect to a dozen SaaS tools securely — what auth infrastructure should I start with? | Your brand was cited | Your brand and a competitor were cited | A competitor was cited | A competitor was cited | Your brand was cited | Neither your brand nor a competitor was cited |
What's the quickest way to add OAuth authentication to AI agents that need to access third-party APIs on behalf of users? | A competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited | A competitor was cited | A competitor was cited | A competitor was cited |
Looking for a drop-in SDK to handle agent-to-API authentication in a TypeScript codebase — what are my options? | Neither your brand nor a competitor was cited | A competitor was cited | A competitor was cited | A competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited |
What tools let me add delegated auth to AI agents without requiring each end user to re-authenticate for every tool call? | Your brand and a competitor were cited | Your brand and a competitor were cited | A competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited | Neither your brand nor a competitor was cited |
Which platforms make it easiest to set up secure auth for MCP servers without building custom OAuth flows from scratch? | Neither your brand nor a competitor was cited | A competitor was cited | A competitor was cited | A competitor was cited | A competitor was cited | Neither your brand nor a competitor was cited |
Turn this matrix into daily prompt monitoring.
Track prompt changesVertical Ranking
| # | Brand | PresencePres. | Share of VoiceSoV | DocsDocs | BlogBlog | MentionsMent. | Avg PosPos | Sentiment |
|---|---|---|---|---|---|---|---|---|
| 1 | WorkOS | 26.7% | 19.5% | 0.7% | 26.0% | 32.0% | #9.9 | +0.39 |
| 2 | Arcade.dev | 24.7% | 19.8% | 0.0% | 24.7% | 18.7% | #10.3 | +0.38 |
| 3 | Nango | 20.7% | 14.9% | 0.0% | 18.7% | 26.0% | #8.6 | +0.34 |
| 4 | Auth0 | 18.7% | 14.9% | 2.7% | 10.0% | 0.0% | #11.5 | +0.42 |
| 5 | Composio | 18.7% | 15.3% | 2.0% | 0.0% | 22.0% | #15.7 | +0.38 |
| 6 | Stytch | 14.0% | 9.1% | 2.7% | 12.0% | 11.3% | #14.8 | +0.37 |
| 7 | Descope | 6.0% | 2.9% | 0.7% | 5.3% | 3.3% | #12.9 | +0.41 |
| 8 | Merge | 4.7% | 2.6% | 0.0% | 4.7% | 12.0% | #13.1 | +0.27 |
| 9 | Oasis Security | 1.3% | 0.6% | 0.0% | 0.0% | 1.3% | #3.0 | +0.10 |
| 10 | Better Auth | 0.7% | 0.3% | 0.0% | 0.0% | 0.7% | #9.0 | +0.40 |
| 11 | Astrix Security | 0.0% | 0.0% | 0.0% | 0.0% | 2.7% | — | — |
| 12 | Keycard.ai | 0.0% | 0.0% | 0.0% | 0.0% | 0.7% | — | — |
| 13 | Operant AI | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | — | — |
Turn this into your team dashboard
Sign up to unlock project-level analytics, daily tracking, actionable insights, custom prompt configurations, adoption tracking, AI traffic analytics and more.
Free trial. Setup comes pre-filled from this report.