Google brings MCP tool support to Apigee hybrid for on-premises and multi-cloud deployments
In short: Google has released Apigee hybrid 1.17.0 (September 14, 2026), adding Model Context Protocol (MCP) support that runs entirely inside a customer's own Kubernetes cluster rather than on Google-managed infrastructure. This lets organizations with data-residency or regulatory requirements turn existing OpenAPI-described APIs into AI-agent-callable tools without MCP request traffic leaving their network. The feature reuses existing Apigee identity, security policies and API product controls, and it is an optional, off-by-default capability enabled via a Helm chart upgrade.
This summary was generated automatically by AI from Google Developers's publication. It is our own text, not a copy of the original — facts, figures and quotes belong to the source, linked above and below.
What changed?
- 1Apigee hybrid 1.17.0 (Sept 14, 2026) adds MCP support with the data plane running inside the customer's own Kubernetes cluster
- 2MCP request/response traffic never leaves the customer's cluster; configuration is pulled from the control plane, not pushed into it
- 3Reuses existing apigee-watcher identity — no new Google service account or IAM role needed
- 4Supports Data Residency–enabled organizations on Subscription and Evaluation tiers
- 5Governance (OAuth 2.0, OIDC, API keys, API product-based tool filtering) applies to MCP endpoints automatically
- 6MCP Discovery Proxies auto-ingest into API hub and API insights for tool-level traffic analytics
- 7Enabled via overrides.yaml enableMcpServer: true and a two-step Helm chart upgrade (operator chart first, org chart second)
- 8Supports end-to-end distributed tracing via W3C trace context forwarded through MCP JSON-RPC _meta fields to OpenTelemetry/CloudTrace
Why it matters
Many enterprises cannot move their API runtime to a public cloud control plane due to regulatory or data-sovereignty requirements; this release lets them expose existing governed APIs as AI-agent tools without that tradeoff, which matters for any agent-building team working with regulated or on-premises backends.
What it means for AI agents and contact centers
If your voice AI or contact-center stack integrates with CRM or backend APIs that must stay on-premises or in a specific jurisdiction, this gives a path to expose those APIs as MCP tools for your agents without routing call or customer data through external infrastructure — worth checking against your own data-residency and latency requirements.
🧪 Worth Testing
For teams building tool-calling voice agents with strict data-residency needs, testing whether Apigee hybrid's local MCP data plane meets latency and governance requirements versus hosted MCP servers could be a concrete, practical evaluation.
Sources
- Google DevelopersOfficialPrimary sourceOriginal article →„Announcing MCP support in Apigee hybrid: agentic tools that run in your own cluster“14 Oct 2026, 03:00
- Published by source
- 14 Oct 2026, 03:00
- Found by our system
- 14 Oct 2026, 21:11
- Summary generated
- 14 Oct 2026, 21:12
This article was written by AI from the original source. Facts, numbers and prices come from the source; missing values are marked “Not specified”. Legal notice, copyright and privacy