Debuly
Speakeasy AI Control Plane logo

Speakeasy AI Control Plane

Governance layer between AI agents and company systems that connects MCP servers, filters risky data, scopes access by role and logs every tool call.

Paid

What is Speakeasy AI Control Plane?

Speakeasy AI Control Plane sits between an organization's AI clients and the systems they reach, so security and platform teams can approve, secure and monitor AI use in one place. It connects agents to SaaS tools, APIs and databases through pre-built integrations, and turns internal APIs into managed MCP servers from an OpenAPI spec. Prompts, responses and agent actions are inspected as they pass, with PII and credentials blocked or redacted, prompt injection and shadow tools detected, and alerts sent to Slack, PagerDuty or a SIEM. Access is distributed by team and role through your identity provider such as Okta or Microsoft Entra ID, using OAuth 2.1 with permissions down to the individual tool. Usage dashboards show adoption by team, tool and user, and an audit trail records each call. It works with clients including Claude, ChatGPT, Cursor and Copilot, and the data plane can be self-hosted in your own VPC. Pay as you go is metered: session storage is $0.35 per million tokens, risk scanning is $0.99 per million tokens, and MCP gateway egress is $20 per GiB. Enterprise is priced by custom contract.

Use cases

  • Govern MCP servers across a company
  • Block PII and secrets from AI prompts
  • Provision AI tool access by role
  • Turn an OpenAPI spec into an MCP server
  • Audit every agent tool call
  • Track AI adoption by team
  • Detect prompt injection and shadow tools

Who it's for

  • Security teams
  • Platform engineers
  • IT administrators
  • AI program leads

Why use Speakeasy AI Control Plane

  • One catalog of approved AI tools replaces scattered integrations that each need their own review.
  • Policy written once applies to every connected AI client, including the agents your teams build.
  • Blocking PII and secrets in flight lowers the chance of sensitive data reaching a model.
  • Role-based provisioning through your identity provider shortens the wait for new hires to get tools.
  • Usage and audit records give security and leadership a view of what AI is touching.

Speakeasy AI Control Plane features

  • Pre-built SaaS integrations
  • Managed MCP servers from an OpenAPI spec
  • Real-time inspection of prompts, responses and tool calls
  • PII and credential blocking or redaction
  • Prompt injection and shadow-tool detection
  • OAuth 2.1 with DCR and PKCE
  • Provisioning through Okta, Entra ID and other SAML or OIDC providers
  • Tool-level permissions by team and role
  • Usage analytics and adoption tracking
  • Audit trail of every tool call
  • Self-hosted data plane option

Speakeasy AI Control Plane integrations

  • Salesforce
  • Slack
  • HubSpot
  • Okta
  • Microsoft Entra ID
  • Auth0
  • WorkOS
  • Google Workspace
  • Ping Identity
  • PagerDuty
  • Claude Code
  • Cursor
  • Codex

How to get started with Speakeasy AI Control Plane

  1. Start a free trial and connect your identity provider.
  2. Add pre-built integrations or publish internal APIs as managed MCP servers from an OpenAPI spec.
  3. Set policies for data inspection, access and approved tools by team and role.
  4. Point AI clients such as Claude, Cursor or your own agents at the control plane.
  5. Watch usage, risk alerts and the audit trail, and adjust access as adoption grows.

Speakeasy AI Control Plane limitations

  • Pay as you go charges for storage, risk scanning and gateway egress
  • Inference the platform runs for you is billed at provider cost
  • SSO, full audit trail and a dedicated engineer need an Enterprise contract
  • Observed session tracking comes from instrumented agent platforms

Speakeasy AI Control Plane pricing

Speakeasy AI Control Plane has no free plan. Plans start at $0.35 usage-based, according to the official pricing page on .

Speakeasy AI Control Plane plans

  • Pay as you go

    Usage-based

    no platform fee; starts with a free trial

    • $0.35 per million tokens of agent session storage
    • $0.99 per million tokens scanned for risk
    • $20 per GiB of MCP gateway egress
  • Enterprise

    Custom

    annual contract on tokens under management

    • SSO and full audit trail
    • Dedicated forward-deployed engineer
    • Early feature access
    • Legal and security reviews

Plans as listed on the official pricing page on . Prices can change and may vary by country or exclude tax. Check the official pricing page.

Speakeasy AI Control Plane alternatives

The main Speakeasy AI Control Plane alternatives listed on Debuly are Kong AI Gateway, Portkey, Arcade, Metorial, CloudByte AI and Gram. The table compares pricing and platforms; order is A–Z.

Speakeasy AI Control Plane compared with its alternatives
ToolPricingStarting pricePlatforms
Speakeasy AI Control PlanePaid$0.35, usage-basedWeb, API, Self-hosted
ArcadeFreemium$25/moWeb, API, Self-hosted
CloudByte AIFreemium$50/moWeb, Self-hosted
GramFree trial$0.35, usage-basedWeb, API, CLI
Kong AI GatewayContact sales—Web, API, Self-hosted
MetorialFreemium$250/moWeb, API, Self-hosted
PortkeyFreemium$49/moWeb, API, Self-hosted
  1. Arcade

    Actions runtime that lets AI agents call tools in Gmail, Slack, Salesforce and more under each user's own permissions.

  2. CloudByte AI

    Observability dashboard for teams using AI coding tools, tracking sessions, token spend, commits, security and DORA metrics.

  3. Gram

    Speakeasy's platform for building, hosting and governing MCP servers, with agent session logging, risk scanning and usage-based billing.

  4. Kong AI Gateway

    Gateway and control layer for LLM, MCP and agent traffic, with usage metering, billing and cost limits built in.

  5. Metorial

    Control plane for AI agents with 1,000+ managed MCP integrations, per-user permissions, SSO and audit logs behind one connection URL.

  6. Portkey

    AI gateway with observability, guardrails, caching and prompt management for teams running LLM apps in production.

Speakeasy AI Control Plane FAQ

How much does Speakeasy AI Control Plane cost?

Pay as you go has no platform fee. Agent session storage is $0.35 per million tokens, risk scanning is $0.99 per million tokens, and MCP gateway egress is $20 per GiB. Enterprise pricing is tailored.

Is there a free trial?

Yes. Pay as you go begins with a trial. Before it ends, add a payment method to move onto the usage-based rates.

What does the Enterprise plan add?

Enterprise is an annual contract based on tokens under management. It adds SSO, complete audit trails, support backed by an SLA that includes a forward-deployed engineer, early access to features, and legal and security reviews.

Which AI clients and identity providers work with it?

The product page names Claude, ChatGPT, Cursor and Copilot as supported clients. For sign-in it lists Okta, Microsoft Entra ID and Auth0, as well as Google Workspace, WorkOS and Ping Identity. Any provider that speaks SAML or OIDC also works.

Can it be self-hosted?

Yes. The data plane can be deployed in your own VPC, with data residency options available on request.

What are alternatives to Speakeasy AI Control Plane?

Alternatives include Kong AI Gateway, Cloudflare AI Gateway, Portkey, Lasso Security and Docker MCP Gateway.

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

SourceTypeAccessed
speakeasy.com/product/ai-control-planeofficial
speakeasy.com/pricingpricing
speakeasy.com/docs/ai-control-plane/org-admin/billingpricing
speakeasy.com/enterpriseofficial

Image credits

  • Logo: from speakeasy.com · fetched · Vendor trademark, used only to identify the product; removed on request
  • Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.