Octopus API Gateway - secure, multi-tenant API gateway for microservices on GKE
Octopus is FutureIM's API gateway for Google Kubernetes Engine. Every request passes through an 8-layer pipeline - observability, network limits, throttling, authentication, authorisation, security, transformation and routing - and every project gets its own root path, upstreams and security policy. It is available on Google Cloud Marketplace.
View in MarketplaceRequest a demoWhat an API gateway does
An API gateway is the single entry point to a set of backend services. Clients call one address; the gateway authenticates the call, applies limits, routes it to the right microservice and reports what happened. Without one, every service has to implement authentication, throttling, routing and logging on its own.
The Octopus request pipeline
- Observe - correlation ID, Prometheus metrics and structured logs, created first so everything downstream is traceable.
- Network - hard-cap rate limiting, IP allow/deny rules and payload-size limits.
- Throttle - a leaky-bucket queue smooths bursts and per-plan quotas cap what each consumer gets.
- Auth - OIDC (Okta, Azure AD / Entra ID, Auth0), signed API keys or LDAP.
- Authz - RBAC roles, OAuth scope guards and Open Policy Agent evaluation.
- Security - CORS, WAF pattern matching, mutual TLS and HMAC request signing, per project.
- Transform - header injection, compression and an opt-in distributed response cache.
- Routing - project and route match, weighted upstream selection and proxying, with circuit breakers per upstream.
Key capabilities
- Multi-tenant by design - organisations own multiple projects; each project has its own root path, upstreams and security policy.
- OpenAPI import - create a project from an OpenAPI / Swagger definition.
- MuleSoft migration - import RAML 0.8 / 1.0 (a single file or a ZIP of the API folder) directly, or convert it to OpenAPI first.
- Agentic capabilities - AI agents draft OpenAPI specs from plain English, review cross-project dependency requests and watch the audit trail for anomalies (details below).
- Security and compliance - Google Cloud Armor OWASP ruleset on by default, RBAC with just-in-time privileged elevation, audit trail and optional CMEK through Cloud KMS.
- API versioning - one project per version, or successive versions with the X-API-Version header, a compatibility check and promotion.
- Operations dashboard and Admin API - overview, traffic, observability, throttling, limits, cache, billing plans, API keys, plugins and deployment views, plus a REST admin API and Prometheus metrics.
Agentic capabilities
Octopus builds AI agents into API governance, so routine review work is done for you and people approve the outcome:
- AIDD (AI-driven design) - drafts an OpenAPI specification from a plain-English description of your API.
- Approval Agent - reviews cross-project dependency requests, so one team cannot quietly depend on another team's API.
- Audit Agent - runs Gemini anomaly detection over the audit trail to surface unusual activity.
Built for Google Cloud
Octopus runs on GKE with Istio behind a Global HTTPS Load Balancer, uses Firestore for configuration and audit data and Memorystore for cache and rate limits. It is delivered through Google Cloud Marketplace, so purchase and billing go through your Google Cloud account.
Octopus API Gateway pricing
Octopus is billed monthly by usage. There is no per-request fee: the usage fee is metered against the real vCPU and memory of the running Octopus pod or pods.
- USD 0.30 per vCPU-hour
- USD 0.06 per GiB-hour of pod memory
| Example pod | CPU | Memory | Per hour | Per month (730 h, 1 pod) |
|---|---|---|---|---|
| Small | 0.5 vCPU | 1 GiB | USD 0.21 | USD 153.30 |
| Medium | 1 vCPU | 2 GiB | USD 0.42 | USD 306.60 |
| Large | 2 vCPU | 4 GiB | USD 0.84 | USD 613.20 |
Examples are arithmetic on the published rates for one pod running all month. Multiply by the number of pods you run. The Octopus usage fee is separate from the Google Cloud services Octopus uses (for example GKE, load balancing, Firestore, Memorystore and Cloud Armor); Google bills those to your account at its own rates.
Octopus vs Apigee
Apigee is Google Cloud's API management product, delivered as a Google-managed service and priced by edition, API call volume and environments. Octopus runs in your own GKE cluster and is billed by the compute it uses, with no per-request fee.
- Choose Apigee when you want a fully Google-managed service, rely on existing Apigee proxies and tooling, or need Apigee-specific capabilities.
- Choose Octopus when you want a gateway in your own GKE cluster with an OIDC / RBAC / WAF / mTLS / HMAC security pipeline, multi-tenant projects, OpenAPI and MuleSoft RAML import and agentic governance - and want cost to follow the capacity you run, not the number of calls.
To compare for your workload: estimate monthly calls and peak requests per second, get an Apigee quote for that volume, size the Octopus pod(s) and apply the rates above, then add the Google Cloud services Octopus uses, which Google bills separately. See the full Octopus vs Apigee comparison.
Get Octopus or see a demo
Purchase Octopus on Google Cloud Marketplace. To see a demo first, email sales@futureim.in or use the WhatsApp option in the Contact section of futureim.com.
Frequently asked questions
What is an API gateway?
An API gateway is a single entry point for client requests to your backend services. It handles routing, authentication, rate limiting, load balancing and monitoring so that individual microservices do not have to.
What is Octopus API Gateway?
Octopus is FutureIM's multi-tenant API gateway for Google Kubernetes Engine. It runs every request through an 8-layer pipeline (observe, network, throttle, auth, authz, security, transform, routing) and adds OpenAPI and MuleSoft RAML import and AI-assisted governance.
Where can I buy Octopus API Gateway?
On Google Cloud Marketplace: the Octopus API Gateway listing from FutureIM. Billing goes through your Google Cloud account.
How much does Octopus API Gateway cost?
Billed monthly by usage: USD 0.30 per vCPU-hour and USD 0.06 per GiB-hour, metered on the running Octopus pod or pods. There is no per-request fee. One pod with 1 vCPU and 2 GiB running all month is about USD 306.60 at those rates.
What else will I pay for besides the Octopus usage fee?
The Google Cloud services Octopus runs on - for example GKE, load balancing, Firestore, Memorystore and Cloud Armor - are billed separately by Google to your account at Google's own rates. The Octopus usage fee covers the software only.
Which authentication methods does Octopus support?
OIDC (Okta, Azure AD / Entra ID, Auth0), signed API keys and LDAP, with RBAC, OAuth scope guards and Open Policy Agent for authorisation.
Can I migrate from MuleSoft to Octopus?
Octopus can import Mule API definitions in RAML 0.8 or 1.0 - a single file or a ZIP of the API folder - either directly into a project or by converting them to OpenAPI first.
Can I create routes from an OpenAPI file?
Yes. Octopus imports OpenAPI / Swagger definitions into projects, and its AIDD assistant can draft an OpenAPI specification from a plain-English description.
How do I get a demo of Octopus?
Email sales@futureim.in or use the WhatsApp option in the Contact section of futureim.com.