Softwr

Technology · head to head

Envoy vs Garden

Envoy logo

Envoy

Technology

A high-performance L7 proxy written in C++ that is configured by an API rather than a config file, and is usually deployed under a control plane.

From
Free
Free plan
Yes
Garden logo

Garden

Developer Tools

Kubernetes development and testing automation with shared caching, now owned by Incredibuild

From
Free
Free plan
Yes

The short version

  • Each has a real cost: Envoy the configuration surface is very large and hand-written bootstrap YAML runs to hundreds of lines for routing that Nginx expresses in twenty, which is why nearly every production deployment sits under a control plane and inherits that control plane's constraints as well.; Garden incredibuild acquired Garden in November 2024 and the pricing page now redirects to documentation, so there is no public commercial offer and no published roadmap for a team betting a platform on it.
  • They diverge on capability: Envoy covers xDS dynamic configuration, Garden covers Action graph.
  • Prices and features above were last checked on 31 August 2026.

Where they differ

Only the attributes on which Envoy and Garden actually diverge.

Attributes where Envoy and Garden differ
AttributeEnvoyGarden
Pricing modelopen-sourceOpen source, no licence fee
PlatformsWebLinux, macOS, Windows, Kubernetes
CategoryTechnologyDeveloper Tools

Identical on both: starting price (Free), free tier (Yes), user rating (Not yet rated).

What each one covers

Drawn from each product's published feature list. An absence here means we hold no record of it - not that the product lacks it.

Only in Envoy

  • xDS dynamic configuration
  • Protocol breadth
  • Filter chain architecture
  • Observability by default
  • Outlier detection
  • Traffic shaping
  • mTLS termination and origination
  • Hot restart

Only in Garden

  • Action graph
  • Shared result caching
  • Remote Kubernetes environments
  • In-cluster builds
  • Same config everywhere
  • Hot reload sync
  • Helm and Kustomize support
  • Templating

What people use each for

The jobs each tool is most often brought in to do.

Envoy

  • Acting as the data plane under a service mesh or Gateway API implementation, which is how the overwhelming majority of deployments use itnot Garden
  • An edge or API gateway that needs per-route retry budgets, circuit breaking and outlier detection rather than round-robin proxyingnot Garden
  • Migrating traffic between service versions or between a monolith and its replacement, using weighted splits and shadow trafficnot Garden
  • Standardising observability across a polyglot estate, so that latency, error rates and tracing look the same regardless of the language a service is written innot Garden

Garden

  • A platform team whose integration test suite takes longer than the coffee break and needs caching to cut the loopnot Envoy
  • An organisation with thirty or more services on Kubernetes where docker-compose no longer approximates productionnot Envoy
  • Giving every developer an ephemeral namespace instead of queuing for one shared staging environmentnot Envoy
  • Running the identical build and test definition locally and in CI so failures reproduce on a laptopnot Envoy

Where each one falls short

Documented limitations, not opinions. Every one is a constraint you would hit in normal use.

Envoy

  • The configuration surface is very large and hand-written bootstrap YAML runs to hundreds of lines for routing that Nginx expresses in twenty, which is why nearly every production deployment sits under a control plane and inherits that control plane's constraints as well.
  • xDS is the real API and it is not stable in the comfortable sense; the v2 API set was removed outright, resource types continue to be deprecated, and your control plane and Envoy binaries have to be upgraded roughly in step or the proxies stop accepting configuration.
  • Extending it properly means writing a C++ filter and building and maintaining your own Envoy binary; the alternatives are Lua, which adds per-request overhead, and proxy-wasm, whose ABI has remained effectively experimental for years with a real performance cost.
  • At sidecar density the per-proxy memory and CPU footprint is a measurable share of cluster capacity, since thousands of workloads each carry a full proxy, and this is precisely the cost that has pushed mesh projects towards node-level or ambient architectures.
  • There is no single vendor selling support for Envoy itself; you get the community plus control-plane vendors such as Solo.io and Tetrate, so an Envoy-level production bug is your own engineers in a C++ codebase unless a support contract happens to cover it.
  • Diagnosing why a request got a particular response involves reading config dumps, the stats endpoint and the RESPONSE_FLAGS codes in access logs rather than a readable error, which is a specific skill you must hire or spend months growing.

Garden

  • Incredibuild acquired Garden in November 2024 and the pricing page now redirects to documentation, so there is no public commercial offer and no published roadmap for a team betting a platform on it.
  • It is tightly bound to Kubernetes, so teams running on ECS, Nomad, plain VMs or serverless get little from it and would need a different tool entirely.
  • The YAML action graph is a real configuration surface: a large monorepo needs someone to own and maintain Garden config as a first-class artefact, which is a standing platform-engineering cost.
  • Shared remote environments mean every developer needs cluster access and a cost allocation model, so cloud spend rises and namespace sprawl becomes a thing you have to police.
  • Caching correctness depends on declaring dependencies accurately; an under-declared action will serve a stale cached pass and hide a genuine test failure, which is a hard class of bug to notice.

Pricing, plan by plan

Envoy

Free

No published plan breakdown. See the Envoy review.

Garden

Free
  • Garden open sourceFree
    • Apache 2.0 licensed CLI
    • Full action graph and caching
    • Self-managed clusters
  • Commercial tier$undefined/year
    • Hosted caching and team features
    • Support and onboarding
    • Pricing no longer published since the Incredibuild acquisition

Which should you pick?

Choose Envoy if

  • You need xds dynamic configuration.
  • You want to start without paying.
  • You also want protocol breadth.

Choose Garden if

  • You need action graph.
  • You want to start without paying.
  • You work on Linux, macOS, Windows, Kubernetes.
  • You also want shared result caching.

Questions people ask

Is Envoy or Garden better?
Neither clearly leads. Envoy starts at Free and Garden at Free, and user ratings are close enough to be indistinguishable. Choose on capability and platform support.
Which is cheaper, Envoy or Garden?
Envoy starts at Free and Garden at Free.
Does Envoy or Garden run on more platforms?
Envoy runs on Web. Garden runs on Linux, macOS, Windows, Kubernetes.
Can I use Envoy for free?
Both have a free tier, so you can try either at no cost before committing.
What is Envoy best used for?
Envoy is most often used for acting as the data plane under a service mesh or gateway api implementation, which is how the overwhelming majority of deployments use it, an edge or api gateway that needs per-route retry budgets, circuit breaking and outlier detection rather than round-robin proxying, migrating traffic between service versions or between a monolith and its replacement, using weighted splits and shadow traffic, standardising observability across a polyglot estate, so that latency, error rates and tracing look the same regardless of the language a service is written in. Of those, acting as the data plane under a service mesh or gateway api implementation, which is how the overwhelming majority of deployments use it and an edge or api gateway that needs per-route retry budgets, circuit breaking and outlier detection rather than round-robin proxying are not what Garden is typically brought in for.
What can Envoy do that Garden cannot?
Envoy covers xDS dynamic configuration, Protocol breadth, Filter chain architecture, Observability by default. Garden covers Action graph, Shared result caching, Remote Kubernetes environments, In-cluster builds.

Answered from the vendors’ own pages

Envoy: Should I run Envoy on its own, or under a control plane?

Almost always under one. Directly authoring xDS or static bootstrap configuration is viable for a handful of routes and becomes unmanageable beyond that. Envoy Gateway, Istio, Contour, Gloo and Consul all exist to generate that configuration for you.

Garden: Is Garden still maintained?

The open source project remains on GitHub under Incredibuild ownership, but the commercial pricing page redirects to the docs and there is no public paid offer.

Envoy: How does it compare with Nginx or HAProxy?

Envoy is dynamically configured over an API and instrumented far more heavily; Nginx and HAProxy are faster to configure and lighter for straightforward reverse proxying. If you never need to change routing without a reload, Envoy is more machinery than the problem requires.

Garden: Do I need Kubernetes?

Effectively yes. Garden is built around Kubernetes environments and offers little to teams on other runtimes.

Envoy: What does it cost?

Nothing to licence; it is Apache 2.0 and there is no paid edition. The cost is engineering time and, for most organisations, a commercial control plane or cloud service that packages it.

Garden: Does it replace my CI system?

No. It runs inside GitHub Actions, GitLab CI or similar and speeds up what those pipelines execute.

Envoy: Can I write extensions without C++?

You can write Lua filters or proxy-wasm modules in Rust, Go, C++ or AssemblyScript. Both carry per-request overhead compared with a native filter, and the Wasm path has been slower to stabilise than the project originally projected.

Garden: What is the actual saving?

It comes from skipping builds and tests whose inputs have not changed, so the benefit scales with how many services you have and how often only a few change.

Envoy: Is it a CNCF project?

Yes, it is a graduated CNCF project licensed under Apache 2.0, which means the trademark and governance sit with the foundation rather than with Lyft or any vendor.

Share

Related pages

Other head to heads