Arcade vs Paragon (2026)

More than 8,000 permission aware tools are available, plus custom or third party MCP servers that the customer builds or brings. Authorization is delegated, so the agent acts as its user within that user's scope, and Arcade handles OAuth across dozens of auth providers. Hosted remote MCP servers connect through Arcade to systems such as Salesforce, ServiceNow, GitHub, Atlassian and Snowflake.

Tool Recommendation, added in October 2026, lets an agent describe a task and get back the few tools that fit instead of loading the whole catalog, which Arcade puts at up to 91% fewer input tokens per session, and a Private Registry lets users browse the tools behind a gateway and request missing ones.

Sourcearcade.devread 2026-10-24

ActionKit is one API and MCP server giving an agent hundreds of integration actions across Paragon's connectors, with a uniform interface, agent optimized action schemas and real time execution. The Connect Portal handles OAuth and API key authorization per end user, so each action runs as that user and agents take authenticated action in the user's own SaaS tools.

Sourcedocs.useparagon.com/learn/actionkitread 2026-09-22

Arcade executes the tool calls an agent makes and returns the results. Sequencing, branching and handoffs between agents live in the customer's own framework. Setup guides cover frameworks such as LangChain, CrewAI, OpenAI Agents, Google ADK and Mastra, among others. No workflow runtime of Arcade's own is named, so retries, routing and deterministic workflow steps sit outside the platform.

Sourcearcade.devread 2026-10-24

Workflows is Paragon's embedded iPaaS for event driven automations, with integration logic authored in TypeScript or a low code editor. A workflow can start from the customer's app, third party webhooks, an integration enabled event, a scheduler or another workflow, and API resources supply reusable triggers and steps. Steps are sequenced, branched and routed deterministically, executions can be reviewed step by step in Monitoring, and CI/CD and Git repo sync are supported.

Sourcedocs.useparagon.com/learn/workflowsread 2026-09-22

Tools in Arcade run when an agent or MCP client calls them. No schedule, event, webhook or inbound queue that starts an agent run is named.

Sourcearcade.devread 2026-10-24

Third party events are forwarded to the customer's agent so it can work in the background and respond to what happens in a user's tools. One API subscribes to any user integration event, with normalized webhook policies across integrations. Workflows can be started by app events, integration webhooks, an API resource webhook or a scheduler, so work begins without a person asking.

Sourceuseparagon.comread 2026-09-22

Agents get authenticated tools that read from and act in outside systems. Arcade names no index, graph, embeddings layer or other retrieval structure that it maintains over the customer's documents.

Sourcearcade.devread 2026-10-24

Managed Sync runs fully managed ingestion pipelines over each end user's third-party sources (files, CRM records, tickets, accounting data). It has automatic backfill, incremental sync, failure recovery and normalized schemas, and scales to hundreds of millions of records. A Permissions API enforces the source system's access control on the ingested data for RAG, and records and file content can then be queried through the Sync API. The synced store persists, scales past any context window and stays queryable.

Sourcedocs.useparagon.com/learn/managed-syncread 2026-09-22

Each user's OAuth tokens and authorization grants are kept, so tools can act without asking for consent again. Those stored credentials carry no agent state. Arcade names no memory that keeps an agent's state or a user's context across runs.

Sourcearcade.devread 2026-10-24

Each connected user's authorization and synced third party records persist. Managed Sync, ActionKit and Workflows have no memory layer that keeps an agent's state or a user's context across runs. The synced data store holds records, not agent memory with a set scope or lifetime.

Sourcedocs.useparagon.com/learn/managed-syncread 2026-09-22

Contextual access runs the customer's own policy logic inline. An access hook decides which tools each user can see. A pre execution hook allows, denies or modifies inputs before each tool call, and a post execution hook allows, denies or modifies the output, so sensitive actions can be blocked and personal data redacted.

Users authorize each scope before a tool acts for them, an agent can line up every authorization a job needs before it starts, and rate limits apply. A tool author can ask the user for input during a call through MCP elicitation, but Arcade names no step where a person reviews or approves an action before it commits.

Sourcedocs.arcade.dev/en/operate/governance/contextual-accessread 2026-10-24

End users authorize each integration for themselves in the Connect Portal and can be shown or particular workflows through Workflow Permissions, and the customer controls which actions its product exposes. Those settings limit what an agent may do, but there is no step where a person reviews, approves or validates an agent action before it commits.

Sourcedocs.useparagon.com/connect-portal/workflow-permissionsread 2026-09-22

Authentication runs against the customer's identity provider, with Okta, Microsoft Entra ID, Auth0 and Stytch as user sources. Authorization is delegated, so an agent acts as its user and never past that user's scope. Contextual access hooks enforce the customer's policies on every tool call. Role based access control has organization admin, project admin and member roles, and SSO and audit logs are included on Enterprise. No security attestation is published.

Sourcearcade.devread 2026-10-24

Compliance covers SOC 2 Type II, HIPAA and GDPR, and the plans include GDPR and SOC 2 compliance as a feature. A Trust Center and security documentation cover encryption standards, access control and a signable DPA. On the access side there is SAML based SSO on Enterprise, managed OAuth and API key authorization per end user through the Connect Portal with signed user tokens, and workflow permissions restricting which users or groups see which integrations.

Sourceuseparagon.com/pricingread 2026-09-22

Every action taken through Arcade leaves one record naming the agent, the user and the system, streamed to the customer's SIEM, and execution tool logs show which tools ran, who they ran for and whether they succeeded. Administrative actions go to a separate audit log that is on by default, with OpenTelemetry export, a filterable dashboard and a REST API for pulling events into the customer's SIEM or compliance tooling.

Sourcedocs.arcade.dev/en/operate/governance/audit-logsread 2026-10-24

Monitoring gives per-event visibility across the integration stack. It keeps searchable atomic logs of every project event, ActionKit tool-call logs with payloads inspected in real time, trigger subscription events and payloads, workflow executions with step-by-step breakdowns and failure details, and a Connected Users dashboard of each user's connection health. Events route to the customer's own stack through Event Destinations (Slack, Datadog, New Relic, Sentry) and the Task History API.

Sourcedocs.useparagon.com/monitoring/overviewread 2026-09-22

Customers choose between Arcade Cloud and their own infrastructure. A turnkey platform that Arcade operates can be deployed into the customer's cloud account, as an Azure Marketplace managed app or through an AWS private offer. A self managed Helm deployment can run on the customer's Kubernetes cluster, for data residency, network isolation or compliance. Enterprise adds VPC and fully air gapped deployment.

Sourcedocs.arcade.dev/en/operate/deployread 2026-10-24

Hosting runs either in Paragon's own cloud or entirely on the customer's infrastructure. On premises hosting comes with a managed instance configuration, and the Enterprise plan includes self hosting and forward deployment.

Sourcedocs.useparagon.com/on-premise/hosting-paragon-on-premiseread 2026-09-22

The catalog holds tools and MCP servers that give an agent actions to take. The sample apps are example code. Arcade names no ready made workflows, templates or prebuilt agents of its own.

Sourcearcade.devread 2026-10-24

The connector catalog and action library give an agent actions to call, but these are integration capabilities, not packaged assets. The use case library offers implementation guides, not assets a customer adopts. Paragon has no ready made workflows, templates or role specific agents of its own.

Sourceuseparagon.com/use-case/libraryread 2026-09-22

Any client, model or framework the customer picks can use Arcade's tools. Setup guides cover the customer's own LLM and frameworks such as LangChain, OpenAI Agents, Google ADK, CrewAI, Mastra and Vercel AI SDK, among others. The evaluation command runs against whichever providers and models the customer names.

Sourcearcade.devread 2026-10-24

Paragon serves integration actions to whatever model or framework the customer runs. ActionKit is delivered as an API or MCP server with agent optimized schemas and is used from the customer's own agent code. Paragon has no model of its own, so the model is the customer's choice throughout.

Sourceuseparagon.comread 2026-09-22

Developers get an API, described in the Arcade API Reference, and a Contextual Access Webhook API. SDKs appear throughout the framework guides, alongside the Arcade CLI and an open source MCP framework in Python and JavaScript for building MCP servers that Arcade hosts or connects. The Arcade plugin, released in October 2026, installs into Claude Code, Cursor, Codex, VS Code, Copilot CLI and Claude Desktop, and any remote MCP client can connect through one server URL. The REST API behind audit logs is called with an Arcade API key.

Sourcedocs.arcade.dev/llms.txtread 2026-10-24

The platform is built for developers. Its APIs and SDKs include a front end Connect SDK, headless Connect Portal APIs, the ActionKit and Sync APIs with published OpenAPI specs, a Permissions API and a Task History API. Builders also get a custom integration builder, custom actions, a hosted demo environment and a self hostable MCP server.

Sourcedocs.useparagon.com/llms.txtread 2026-09-22

The arcade evals command runs evaluation suites that the customer writes against its tools. It gives detailed results and critic feedback, a failures only view, multiple runs for stability and comparisons across providers and models in one run. Capture mode records the tool calls a model makes without scoring them, so they can become test expectations, and Arcade describes its evals as catching hallucinated and destructive calls before they run.

Sourcedocs.arcade.dev/en/build/create-tools/evaluate-tools/run-evaluationsread 2026-10-24

Changes move through Release Environments, which test and deploy them before production, and versioned releases name breaking changes. A hosted demo environment and an ActionKit playground let teams try actions, and logs help debug failures. There is no way to score a change, such as test cases, a pass rate or a judge verdict.

Sourcedocs.useparagon.com/deploying-integrations/release-environmentsread 2026-09-22

Actions go through authenticated APIs and MCP tools. Arcade names no browser, desktop or remote computer session that it runs for an agent. Some toolkits connect to scraping and sandbox services that other vendors run.

Sourcearcade.devread 2026-10-24

Paragon reaches third party apps through their APIs and its own connectors. It runs no browser, desktop or remote computer session for an agent, and no headless fetch or third party browser engine is wired in.

Sourcedocs.useparagon.com/learn/actionkitread 2026-09-22