Speakeasy
An enterprise AI control plane for connecting, securing, distributing, and observing MCPs, skills, assistants, and agent sessions.
About Speakeasy
Speakeasy is an enterprise AI control plane for teams that need to connect agents to internal APIs, SaaS tools, and data while keeping access governed. It centralizes MCP servers, skills, and assistants, then applies fine-grained permissions, authentication, runtime controls, and audit visibility to tool calls. The platform can generate SDKs and MCP servers from OpenAPI definitions, curate focused toolsets, manage public or private servers, and publish generated packages. Its gateway gives clients such as Claude, Cursor, ChatGPT, Copilot, Codex, and internal agents one managed entry point. Teams can provision access by role, inspect prompts, responses, and calls, attribute usage costs, set budgets, and block risky activity. The combined workflow suits organizations standardizing agent access across departments, while smaller developers can begin with the free generation tier. Teams should plan for usage accounting, toolset design, and an enterprise conversation when they need advanced security, service commitments, or organization-wide governance.
Best for
Best use cases
Generate an SDK or MCP server and publish the generated package.
Give multiple AI clients governed access through a single MCP gateway.
Control team permissions, monitor tool calls, and manage AI usage costs.
Compare before choosing
The free tier is useful for limited generation work, but it includes one language and 250 operations per language.
Enterprise adds SSO, audit logs, custom domains, priority support, and SLAs for governed deployments.
Request-based MCP usage can grow faster when a custom tool orchestrates several sub-tools, because every tool call counts.
Features
Generate SDKs and MCP servers
Turn OpenAPI definitions into SDKs or MCP servers, then publish generated packages through the supported toolchain.
Govern agent access
Route AI clients through one gateway with authentication, permissions, server controls, and an audit trail for tool calls.
Manage organizational AI usage
Provision access by team, inspect interactions, attribute token usage, set budgets, and cap runaway costs.
Pros
- A $0 plan lets developers try SDK and MCP generation before committing to an enterprise contract.
- One gateway can connect several AI clients to curated MCP servers with centralized permissions and audit visibility.
- OpenAPI generation and package publishing reduce repetitive integration work for API teams.
Cons
- The free plan is limited to one language and 250 operations per language, which may be restrictive for larger APIs.
- Enterprise pricing requires a sales conversation, making advanced governance costs harder to estimate upfront.
- Every orchestrated sub-tool call counts as a request, so complex custom tools can increase measured usage.
Pricing Plans
Free
- SDK and MCP generation
- Automatic publishing to a package manager
- Access to the OpenAPI toolchain
- 250 operations per language
- Community Slack support
- License for 1 language
Enterprise
- OpenAPI import and custom tool creation
- Multi-tool workflow builder and LLM playground testing
- OAuth options and server authentication
- Public and private server control with custom domains
- SSO and audit logs
- Priority Slack, Teams, chat, and email support
- Service and support SLAs
- Concierge onboarding and legal or security reviews
* Pricing information is for reference only. Please refer to the latest announcement on Speakeasy's official website.