diff --git a/.github/workflows/documenter.yml b/.github/workflows/documenter.yml
index 7694b463..5b8a639a 100644
--- a/.github/workflows/documenter.yml
+++ b/.github/workflows/documenter.yml
@@ -30,6 +30,12 @@ jobs:
           key: ${{ runner.os }}-${{ env.cache-name }}-${{ hashFiles('docs/CondaPkg.toml') }}
           restore-keys: |
             ${{ runner.os }}-${{ env.cache-name }}-
+      - name: Cache CmdStan
+        id: cache-cmdstan
+        uses: actions/cache@v2
+        with:
+          path: ${{ env.CMDSTAN_PATH }}
+          key: cmdstan-${{ env.CMDSTAN_VERSION }}-${{ runner.os }}
       - name: Download and build CmdStan
         if: steps.cache-cmdstan.outputs.cache-hit != 'true'
         run: |
diff --git a/docs/Project.toml b/docs/Project.toml
index f7aeb2ca..66b15b75 100644
--- a/docs/Project.toml
+++ b/docs/Project.toml
@@ -34,7 +34,7 @@ CairoMakie = "0.8.9, 0.9, 0.10"
 DataFrames = "1"
 DimensionalData = "0.23, 0.24"
 Distributions = "0.25"
-Documenter = "0.27"
+Documenter = "1"
 EvoTrees = "0.14.8, 0.15, 0.16"
 MCMCDiagnosticTools = "0.3"
 MLJBase = "1"
diff --git a/docs/make.jl b/docs/make.jl
index 445d662b..63a0671a 100644
--- a/docs/make.jl
+++ b/docs/make.jl
@@ -67,6 +67,14 @@ doctestfilters = [
     r"\s+\"created_at\" => .*",  # ignore timestamps in doctests
 ]
 
+prettyurls = haskey(ENV, "CI")
+
+warnonly = [
+    :missing_docs,  # we don't reexport or document all of MCMCDiagnosticTools
+    :cross_references,  # we reference the pages built by PlutoStaticHTML with pretty relative URLs
+    :linkcheck,  # avoid checking links to pages built with PlutoStaticHTML
+]
+
 makedocs(;
     modules,
     sitename="ArviZ.jl",
@@ -90,16 +98,16 @@ makedocs(;
     ],
     checkdocs=:exports,
     format=Documenter.HTML(;
-        prettyurls=haskey(ENV, "CI"),
+        prettyurls,
+        size_threshold=2 * 10^6,  # 2Mb, needed since PlutoStaticHTML embeds images in markdown
         assets=["assets/favicon.ico", "assets/custom.css"],
         sidebar_sitename=false,
         canonical="stable",
+        analytics="G-W1G68W77YV",
     ),
     doctestfilters,
     linkcheck=true,
-    analytics="G-W1G68W77YV",
-    # allow linkcheck to fail so we can use pretty links to PlutoStaticHTML notebooks
-    strict=Documenter.except(:footnote, :linkcheck, :missing_docs),
+    warnonly,
 )
 
 deploydocs(; repo="github.com/arviz-devs/ArviZ.jl.git", devbranch="main", push_preview=true)
diff --git a/docs/src/index.md b/docs/src/index.md
index d2882954..8a8591db 100644
--- a/docs/src/index.md
+++ b/docs/src/index.md
@@ -4,14 +4,14 @@ ArviZ.jl is a Julia meta-package for exploratory analysis of Bayesian models.
 It is part of the [ArviZ project](https://www.arviz.org/), which also includes a related [Python package](https://python.arviz.org/).
 
 ArviZ consists of and re-exports the following subpackages, along with extensions integrating them with InferenceObjects:
-- [InferenceObjects.jl](https://julia.arviz.org/InferenceObjects): a base package implementing the [`InferenceData`](@ref) type with utilities for building, saving, and working with it
-- [MCMCDiagnosticTools.jl](https://julia.arviz.org/MCMCDiagnosticTools): diagnostics for Markov Chain Monte Carlo methods
-- [PSIS.jl](https://julia.arviz.org/PSIS): Pareto-smoothed importance sampling
-- [PosteriorStats.jl](https://julia.arviz.org/PosteriorStats): common statistical analyses for the Bayesian workflow
+- [InferenceObjects.jl](https://julia.arviz.org/InferenceObjects/): a base package implementing the [`InferenceData`](@ref) type with utilities for building, saving, and working with it
+- [MCMCDiagnosticTools.jl](https://julia.arviz.org/MCMCDiagnosticTools/): diagnostics for Markov Chain Monte Carlo methods
+- [PSIS.jl](https://julia.arviz.org/PSIS/): Pareto-smoothed importance sampling
+- [PosteriorStats.jl](https://julia.arviz.org/PosteriorStats/): common statistical analyses for the Bayesian workflow
 
 Additional functionality can be loaded with the following packages:
-- [ArviZExampleData.jl](https://julia.arviz.org/ArviZExampleData): example `InferenceData` objects, useful for demonstration and testing
-- [ArviZPythonPlots.jl](https://julia.arviz.org/ArviZPythonPlots): Python ArviZ's library of plotting functions for Julia types
+- [ArviZExampleData.jl](https://julia.arviz.org/ArviZExampleData/): example `InferenceData` objects, useful for demonstration and testing
+- [ArviZPythonPlots.jl](https://julia.arviz.org/ArviZPythonPlots/): Python ArviZ's library of plotting functions for Julia types
 
 See the navigation bar for more useful packages.