[Docs] Fix outdated facts, broken links, and examples

README.md

@@ -24,7 +24,7 @@ It streamlines development, training, and inference, and is compatible with any
 
 ## Latest news ✨
 - [2026/08] [dstack 0.21.0: Pydantic v2, Gateway replicas](https://github.com/dstackai/dstack/releases/tag/0.21.0)
-- [2026/07] [dstack 0.20.29: Presets — agent-driven inference optimization (experimental)](https://dstack.ai/docs/concepts/presets/)
+- [2026/07] [dstack 0.20.29: Presets — agent-driven inference optimization](https://dstack.ai/docs/concepts/presets/)
 - [2026/07] [dstack 0.20.27: Slurm backend](https://github.com/dstackai/dstack/releases/tag/0.20.27)
 - [2026/05] [dstack 0.20.21: Kubernetes multiple clusters](https://github.com/dstackai/dstack/releases/tag/0.20.21)
 - [2026/05] [dstack 0.20.20: NVIDIA Dynamo integration](https://github.com/dstackai/dstack/releases/tag/0.20.20)
@@ -107,7 +107,7 @@ AI agents like Claude, Codex, and Cursor can now create and manage fleets and su
 * [Dev environments](https://dstack.ai/docs/concepts/dev-environments) — Launch dev environments to be accessed by agents or from your IDE
 * [Tasks](https://dstack.ai/docs/concepts/tasks) — Run training, batch or other jobs across a single node or clusters
 * [Services](https://dstack.ai/docs/concepts/services) — Deploy model inference as secure and scalable endpoints
-* [Presets](https://dstack.ai/docs/concepts/presets) — Agent-driven inference optimization (experimental)
+* [Presets](https://dstack.ai/docs/concepts/presets) — Agent-driven inference optimization
 * [Volumes](https://dstack.ai/docs/concepts/volumes) — Managing instance and network volumes for persisting data
 
 Configuration can be defined as YAML files within your repo.

frontend/src/pages/Project/Details/Settings/constants.tsx

@@ -14,7 +14,7 @@ export const CLI_INFO = {
             </p>
             <p>
                 To learn how to install the CLI, refer to the{' '}
-                <a href={'https://dstack.ai/docs/installation#set-up-the-cli'} target="_blank">
+                <a href={'https://dstack.ai/docs/installation#cli'} target="_blank">
                     installation
                 </a>{' '}
                 guide.

mkdocs.yml

@@ -107,12 +107,19 @@ plugins:
         "docs/reference/cli/dstack/endpoint.md": "docs/reference/cli/dstack/preset.md"
         "docs/examples/llms/llama31.md": "docs/examples/inference/vllm.md"
         "docs/examples/llms/llama32.md": "docs/examples/inference/vllm.md"
-        "docs/examples/llms/qwen36.md": "docs/examples/models/qwen36.md"
+        "docs/examples/llms/qwen36.md": "docs/examples/inference/sglang.md"
         "examples/llms/llama31/index.md": "docs/examples/inference/vllm.md"
         "examples/llms/llama32/index.md": "docs/examples/inference/vllm.md"
-        "examples/llms/qwen36/index.md": "docs/examples/models/qwen36.md"
+        "examples/llms/qwen36/index.md": "docs/examples/inference/sglang.md"
         "docs/examples/deployment/nim/index.md": "docs/examples/inference/nim.md"
         "docs/examples/deployment/vllm/index.md": "docs/examples/inference/vllm.md"
+        "docs/examples/models/qwen36.md": "docs/examples/inference/sglang.md"
+        "docs/examples/models/deepseek-v4.md": "docs/examples/inference/sglang.md"
+        "docs/examples/models/wan22/index.md": "docs/examples.md"
+        "docs/examples/accelerators/intel/index.md": "docs/examples.md"
+        "docs/examples/llms/deepseek/index.md": "docs/examples/inference/sglang.md"
+        "docs/examples/llms/llama/index.md": "docs/examples/inference/vllm.md"
+        "docs/examples/misc/docker-compose/index.md": "docs/concepts/dev-environments.md#docker-in-docker"
         "backends.md": "docs/concepts/backends.md"
         "blog/monitoring-gpu-usage.md": "blog/posts/dstack-metrics.md"
         "blog/inactive-dev-environments-auto-shutdown.md": "blog/posts/inactivity-duration.md"
@@ -193,8 +200,8 @@ plugins:
         "examples/inference/vllm.md": "docs/examples/inference/vllm.md"
         "examples/inference/nim.md": "docs/examples/inference/nim.md"
         "examples/inference/trtllm.md": "docs/examples/inference/trtllm.md"
-        "examples/models/deepseek-v4.md": "docs/examples/models/deepseek-v4.md"
-        "examples/models/qwen36.md": "docs/examples/models/qwen36.md"
+        "examples/models/deepseek-v4.md": "docs/examples/inference/sglang.md"
+        "examples/models/qwen36.md": "docs/examples/inference/sglang.md"
         "examples/accelerators/amd.md": "docs/examples/accelerators/amd.md"
         "examples/accelerators/tpu.md": "docs/examples/accelerators/tpu.md"
         "examples/accelerators/tenstorrent.md": "docs/examples/accelerators/tenstorrent.md"
@@ -335,9 +342,6 @@ nav:
             - vLLM: docs/examples/inference/vllm.md
             - NIM: docs/examples/inference/nim.md
             - TensorRT-LLM: docs/examples/inference/trtllm.md
-        - Models:
-            - DeepSeek V4: docs/examples/models/deepseek-v4.md
-            - Qwen 3.6: docs/examples/models/qwen36.md
         - Accelerators:
             - AMD: docs/examples/accelerators/amd.md
             - Tenstorrent: docs/examples/accelerators/tenstorrent.md

mkdocs/.justfile

@@ -6,9 +6,10 @@ default:
     @just --list
 
 # Preview the docs site with live-reload
-mkdocs-serve:
+[arg("port", long="port")]
+mkdocs-serve port="8000":
     # --livereload works around live-reload bugs in recent mkdocs versions
-    uv run mkdocs serve --livereload -s
+    uv run mkdocs serve --livereload -s -a 127.0.0.1:{{port}}
 
 # Build the docs site to ./site
 mkdocs-build:

mkdocs/assets/stylesheets/extra.css

@@ -1980,10 +1980,9 @@ img.border {
 a[href^="http"]:not(:where(
   /* skip if marked with external-skip */
   .external-skip,
-  /* exclude http:// dstack links */
-  [href^="http://dstack.ai"],
-  /* exclude https://dstack.ai links */
-  [href^="https://dstack.ai"],
+  /* exclude dstack.ai links, except non-docs product pages */
+  [href^="http://dstack.ai"]:not([href*="dstack.ai/products"]),
+  [href^="https://dstack.ai"]:not([href*="dstack.ai/products"]),
   /* exclude md-content__button links */
   .md-content__button,
 )):after {

mkdocs/blog/posts/intel-gaudi.md

@@ -158,7 +158,7 @@ $ dstack apply -f examples/single-node-training/trl/intel/.dstack.yml -R
 
 `dstack` will automatically create containers according to the run configuration and execute them across the fleet.
 
-> Explore our [examples](../../docs/examples/accelerators/intel/index.md) to learn how to train and deploy large models on
+> Explore our [examples](../../docs/examples.md) to learn how to train and deploy large models on
 > Intel Gaudi AI Accelerator.
 
 !!! info "Intel Tiber AI Cloud"

mkdocs/docs/concepts/backends.md

@@ -90,7 +90,7 @@ There are two ways to configure AWS: using an access key or using the default cr
                     "ec2:DeletePlacementGroup",
                     "ec2:DeleteVolume",
                     "ec2:DescribeAvailabilityZones",
-                    "ec2:DescribeCapacityReservations"
+                    "ec2:DescribeCapacityReservations",
                     "ec2:DescribeImages",
                     "ec2:DescribeInstances",
                     "ec2:DescribeInstanceAttribute",
@@ -314,7 +314,8 @@ There are two ways to configure Azure: using a client secret or using the defaul
 
     ```shell
     SUBSCRIPTION_ID=...
-    az ad sp create-for-rbac
+    DSTACK_ROLE=...
+    az ad sp create-for-rbac \
         --name dstack-app \
         --role $DSTACK_ROLE \
         --scopes /subscriptions/$SUBSCRIPTION_ID \
@@ -401,6 +402,8 @@ There are two ways to configure Azure: using a client secret or using the defaul
       - name: main
         backends:
           - type: azure
+            subscription_id: 06c82ce3-28ff-4285-a146-c5e981a9d808
+            tenant_id: f84a7584-88e4-4fd2-8e97-623f0a715ee1
             creds:
               type: default
             regions: [westeurope]
@@ -415,6 +418,8 @@ There are two ways to configure Azure: using a client secret or using the defaul
       - name: main
         backends:
           - type: azure
+            subscription_id: 06c82ce3-28ff-4285-a146-c5e981a9d808
+            tenant_id: f84a7584-88e4-4fd2-8e97-623f0a715ee1
             creds:
               type: default
             regions: [westeurope]
@@ -433,6 +438,8 @@ There are two ways to configure Azure: using a client secret or using the defaul
       - name: main
         backends:
           - type: azure
+            subscription_id: 06c82ce3-28ff-4285-a146-c5e981a9d808
+            tenant_id: f84a7584-88e4-4fd2-8e97-623f0a715ee1
             creds:
               type: default
             regions: [westeurope]
@@ -739,6 +746,7 @@ gcloud projects list --format="json(projectId)"
       - name: main
         backends:
           - type: gcp
+            project_id: gcp-project-id
             creds:
               type: default
 
@@ -891,7 +899,7 @@ projects:
 
 <span id="datacrunch"></span>
 
-### Verda (formerly DataCrunch) { #verda }
+### Verda { #verda }
 
 Log into your [Verda](https://console.verda.com/signin) account, click Keys in the sidebar, find `REST API Credentials` area and then click the `Generate Credentials` button.
 
@@ -1399,7 +1407,7 @@ projects:
 
     This applies to offers shown in `dstack apply` (run plans), during provisioning, and in `dstack offer`. Unlike other backends, offers for the `kubernetes` backend always reflect the lower limit of the range.
 
-> To learn more, see the [Lambda](../examples/clusters/lambda/#kubernetes) and [Crusoe](../examples/clusters/crusoe/#kubernetes) examples.
+> To learn more, see the [Lambda](../examples/clusters/lambda.md#kubernetes) and [Crusoe](../examples/clusters/crusoe.md#kubernetes) examples.
 
 ### Slurm
 
@@ -1648,5 +1656,3 @@ projects:
     ```
 
     </div>
-
-Also, the `vastai` backend supports on-demand instances only. Spot instance support coming soon.

mkdocs/docs/concepts/dev-environments.md

@@ -267,13 +267,13 @@ init:
 
 </div>
 
-Cannot be used with `python` or `image`. Not supported on `runpod`, `vastai`, or `kubernetes`.
+Cannot be used with `python` or `image`. Not supported on `daytona`, `lium`, `runpod`, `slurm`, or `vastai`.
 
 #### Privileged mode
 
 To enable privileged mode, set [`privileged`](../reference/dstack.yml/dev-environment.md#privileged) to `true`.
 
-Not supported with `runpod`, `vastai`, and `kubernetes`.
+Not supported with `daytona`, `lium`, `runpod`, `slurm`, and `vastai`.
 
 #### Private registry
 

mkdocs/docs/concepts/events.md

@@ -26,6 +26,8 @@ In the UI, the `Events` tab is included on the run, job, fleet, and instance pag
 
 ![](https://dstack.ai/static-assets/static-assets/images/dstack-ui-events-run.png){ width=800 }
 
+> The global `Events` page with advanced filtering is only available in [dstack Sky](https://dstack.ai/products/sky/) and [dstack Factory](https://dstack.ai/products/factory/).
+
 ## CLI
 
 To query events via the CLI, use the `dstack event` command.

mkdocs/docs/concepts/exports.md

@@ -8,8 +8,7 @@ description: Exporting resources across projects
 Exports allow making resources from one project available to other projects. When a project exports a resource,
 the specified importer projects can see and use it as if it were their own.
 
-!!! info "Export support"
-    Exports are only supported in `dstack` Sky and `dstack` Factory.
+> Exports are only supported in [dstack Sky](https://dstack.ai/products/sky/) and [dstack Factory](https://dstack.ai/products/factory/).
 
 An export is created in the exporter project and specifies the resources to export and the
 importer projects that will gain access to them.
@@ -20,7 +19,7 @@ for running tasks, dev environments, and services. Imported resources appear wit
 
 Currently, [SSH fleets](fleets.md#ssh-fleets) and [gateways](gateways.md) can be exported.
 
-!!! info "Required project role"
+??? info "Required role"
     The user creating or updating an export must have the project admin role on both the exporter project and
     any importer project they add. Alternatively, a global admin can add any project as an importer.
 
@@ -128,10 +127,10 @@ $ dstack export create global-export --gateway shared-gateway --global
 
 ```
 
-Only promoting an export to global requires the global admin role. Regular project admins can add or remove resources, remove global status, or delete the export.
-
 </div>
 
+Only promoting an export to global requires the global admin role. Regular project admins can add or remove resources, remove global status, or delete the export.
+
 ## Access imported resources
 
 From the importer project's perspective, use `dstack import list` (or simply `dstack import`) to list all imports in the project — i.e., all exports from other projects that this project has been granted access to:

mkdocs/docs/concepts/fleets.md

@@ -7,7 +7,7 @@ description: Managing pools of compute instances
 
 Before submitting runs, you must create a fleet. Fleets act as both pools of instances and templates for how those instances are provisioned.
 
-> `dstack` supports two fleet types: [backend fleets](#backend-fleet) (which are provisioned dynamically in the cloud or on Kubernetes), and [SSH fleets](#ssh-fleet) (which use existing on-prem servers).
+> `dstack` supports two fleet types: [backend fleets](#backend-fleets) (which are provisioned dynamically in the cloud or on Kubernetes), and [SSH fleets](#ssh-fleets) (which use existing on-prem servers).
 
 ## Apply a configuration
 
@@ -103,10 +103,10 @@ To create a fleet, define its configuration in a YAML file. The filename must en
     `dstack apply` automatically connects to on-prem servers, installs the required dependencies, and adds them to the created fleet.
 
     ??? info "Host requirements"
-        1. Hosts must be Linux-based and have Docker pre-installed.
+        1. Hosts must be Linux-based with systemd, have a 64-bit x86 or ARM CPU, and have Docker pre-installed.
 
         === "NVIDIA"
-            2. Hosts with NVIDIA GPUs must also be pre-installed with CUDA 12.1 and
+            2. Hosts with NVIDIA GPUs must also be pre-installed with NVIDIA drivers that support CUDA 13.0 and
             [NVIDIA Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html).
 
         === "AMD"
@@ -120,7 +120,7 @@ To create a fleet, define its configuration in a YAML file. The filename must en
 
         3. The user specified must have passwordless `sudo` access.
 
-        4. The SSH server must be running and configured with `AllowTcpForwarding yes` in `/etc/ssh/sshd_config`.
+        4. The SSH server must be running and configured with `AllowTcpForwarding yes` in `/etc/ssh/sshd_config`. The SSH key must be RSA, ECDSA, or Ed25519.
 
         5. The firewall must allow SSH and should forbid any other connections from external networks.
 
@@ -134,7 +134,7 @@ Backend fleets support [many options](../reference/dstack.yml/fleet.md); see som
 
 ### Cluster placement
 
-Both [backend fleets](#backend-fleet) and [SSH fleets](#ssh-fleet) allow the `placement` property to be set to `cluster`. 
+Both [backend fleets](#backend-fleets) and [SSH fleets](#ssh-fleets) allow the `placement` property to be set to `cluster`. 
 
 This property ensures that instances are interconnected. This is required for running [multi-node tasks](tasks.md#nodes).
 
@@ -320,7 +320,10 @@ Backend fleets allow you to set `backends` to specify which backends are allowed
 
 Instances of a backend fleet that aren't running any jobs are `idle` and can be reused by new runs.
 `dstack` terminates idle instances above [`nodes.min`](#nodes) after
-[`idle_duration`](../reference/dstack.yml/fleet.md#idle_duration), which defaults to 3 days.
+[`idle_duration`](../reference/dstack.yml/fleet.md#idle_duration). If `idle_duration` is not set,
+it defaults to 3 days for pre-provisioned instances and to 5 minutes for instances provisioned
+when you submit runs. Instances of [container-based](backends.md#container-based) backends are
+terminated as soon as they become idle.
 
 For `idle_duration` to have an effect, `nodes` must be a range. Set `min` to `0` and `target` to
 the number of instances to pre-provision:
@@ -498,6 +501,34 @@ $ dstack fleet
 
 </div>
 
+### Update fleets
+
+To update a fleet, change its configuration and pass it to `dstack apply`. `dstack` will automatically detect changes and suggest an in-place update.
+
+<div class="termy">
+
+```shell
+$ dstack apply -f fleet.dstack.yml
+
+Found fleet my-fleet. Detected changes that can be updated in-place:
+- Configuration properties:
+  - nodes
+  - max_price
+
+Update the fleet in-place? [y/n]:
+```
+
+</div>
+
+If approved, `dstack` updates the fleet without re-creating it. For example, if `nodes.min` is increased, `dstack` provisions the missing instances, and if `nodes.max` is decreased, `dstack` terminates idle instances above it.
+
+??? info "Supported properties"
+    Backend fleets support in-place changes to `nodes`, `resources`, `backends`, `regions`, `availability_zones`, `instance_types`, `spot_policy`, `max_price`, `reservation`, and `tags`. Except for `nodes`, these changes only apply to instances provisioned after the update.
+
+    SSH fleets support adding and removing `hosts`. Added hosts are provisioned, and removed hosts are terminated.
+
+    Changes to other properties require re-creating the fleet. In this case, `dstack apply` offers to delete the fleet and create it again.
+
 ### Delete fleets
 
 When a fleet isn't used by a run, you can delete it by passing the fleet configuration to `dstack delete`:

mkdocs/docs/concepts/presets.md

@@ -7,8 +7,6 @@ description: Creating and reusing optimized model inference configurations
 
 Presets offer a toolkit that streamlines agent-based model inference optimization and a portable format to deploy the optimized inference endpoint to any cloud, Kubernetes cluster, or on-prem fleet.
 
-> The presets feature is experimental and may change.
-
 ??? info "Prerequisites"
     Before using presets, make sure you’ve [installed](../installation.md) the server and CLI, and created a [fleet](fleets.md).
 
@@ -70,7 +68,7 @@ Create the preset dsv4-flash? [y/n]: y
 
 The command executes entirely locally and uses the locally installed `claude` or `codex` CLI along with `dstack`'s bundled skills (see [Agent](#agent)). The agent uses a `dstack` task to find the best serving configuration for the available fleet offers, then submits it as a `dstack` service for a final benchmark.
 
-You can stop watching with `Ctrl`+`C` at any time. The agent keeps running, and `dstack preset logs -f` follows it again. Resume an interrupted creation with `dstack preset resume`:
+Pressing `Ctrl`+`C` asks whether to stop the creation. If you don't stop it, the agent keeps running, and `dstack preset logs -f` follows it again. Resume an interrupted creation with `dstack preset resume`:
 
 <div class="termy">