Podcast and audio publishing

Host, master, transcribe, and publish podcast and audio content on Quake AI Compute and Object Storage. You operate RSS feed generation, enclosure URLs, and CDN settings; Quake AI provides compute for publishing workers, Object Storage for episode libraries, and networking for public endpoints.

What this is for#

Podcasters and audio publishers who want to own their feed and library run the publishing pipeline themselves instead of renting a hosting SaaS. Quake AI provides compute for mastering, transcoding, and transcription workers, plus object storage for the episode library and RSS document; you operate the feed generation and CDN configuration. The outcome is an owned audio publishing stack with flat egress, deployable from validated OpenTofu templates and their companion tutorials.

Reference architecture#

Download diagram: SVG, PNG, and PDF.

The base is two validated templates: the Audio worker template runs the CPU worker that masters, transcodes, and generates the feed, and the S3 Storage with ACLs template holds the episode library and RSS document. A third-party CDN fronts the bucket for listener delivery. A Simple VM is an alternative worker host when you do not need the audio-worker preset.

1. Listeners and podcast apps. Subscribers fetch the RSS feed and pull episode enclosures over HTTPS, typically through a CDN.

2. Third-party CDN. A CDN such as Cloudflare or Fastly caches episode objects and serves them close to listeners. Quake AI has no native CDN product, so you bring the provider and point it at the bucket as origin.

3. Publishing and transcode worker. A Compute instance runs the feed generator, audio normalization, and optional transcription. The Audio worker template provisions the CPU worker host; a Simple VM is the alternative when you configure the worker yourself.

4. Object Storage library. Masters, normalized MP3s, cover art, and the RSS document live in S3-compatible Object Storage. The S3 Storage with ACLs template provisions the bucket and ACL policy.

Services involved#

Get started#

• Simple VM template and its deploy tutorial: single Compute instance for a feed generator, transcoding worker, or lightweight publishing app.
• Audio worker template and its deploy tutorial: CPU worker for transcode, normalization, and transcription jobs.
• S3 Storage with ACLs template and its deploy tutorial: Object Storage bucket with ACL policy for your episode and asset library.
• Upload objects to Object Storage: store masters, normalized MP3s, and cover art in the bucket your template provisions.
• OpenTofu template library: browse validated IaC starting points for publishing and storage layouts.

Estimate the cost#

Migrating an existing podcast / audio publishing workload?#

Move a podcast or audio publishing stack that already runs on AWS EC2 and S3, another hyperscaler, or a SaaS host. The outcome is the same episode library on Quake AI Object Storage with RSS, enclosure URLs, and CDN configuration cut over in a controlled order.

Follow this cutover path. Each step links an existing migration page; this section composes those pages into a workload-shaped sequence rather than duplicating their steps.

1. Map your source provider. Start with the concept-translation page for your current cloud: Coming from AWS, Coming from Azure, Coming from GCP, Coming from DigitalOcean, or Coming from Hetzner.

2. Stand up the target shape on Quake AI. Pick the greenfield templates that match your publishing layout: Simple VM for a feed or worker VM, and S3 Storage with ACLs for the audio library bucket.

3. Move compute workloads. Rebuild or migrate feed generators, transcode workers, and publishing VMs with Migrate from EC2 (or the matching compute migration page for your source provider).

4. Sync the audio library. Copy episode files, masters, and cover art with Migrate from S3 (or the matching object migration page for your source provider).

5. Update RSS and CDN endpoints. Publish the new feed URL and enclosure base path, configure your CDN or reverse proxy to serve episode objects from the Quake AI bucket, then switch the public hostname when validation passes.

Workload-specific cutover callouts#

• Audio-library bucket sync. Plan a bulk object copy with integrity checks (ETag or checksum validation) before you change enclosure URLs in the feed. Resume interrupted transfers and spot-check recent episodes after sync completes.
• RSS and CDN endpoint update. Quake AI has no native CDN product. Update your CDN provider (Cloudflare, Fastly, or similar) and any dynamic feed generator so <enclosure> URLs resolve to the Quake AI Object Storage path or signed URL pattern you provisioned.
• Feed continuity. Validate the RSS XML after URL changes: every enclosure must return HTTP 200, byte length must match the feed metadata, and podcast apps that cache feeds should see stable episode identifiers across the cutover window.
• Rollback path. Keep the source bucket and feed endpoint available until subscriber playback error rates stabilize. Roll back by restoring the previous feed document and CDN origin configuration if validation or health checks fail.

Considerations and limits#

• You operate publishing and the feed. Quake AI provides compute and storage; RSS generation, enclosure URLs, and transcription are yours under the shared responsibility model.
• No native CDN. Quake AI has no CDN product. Front the episode bucket with a third-party CDN for listener-facing delivery.
• CPU-only compute. Compute is AMD EPYC with no GPU option (compute FAQ). CPU workers fit audio transcode, normalization, and speech-to-text on smaller models.
• Flat egress. Quake AI applies a no-egress-fee policy for outbound transfer, which suits feeds that serve large audio files to many subscribers.
• Three US regions. All current regions are in the United States. Global delivery relies on the CDN layer in front of the origin.
• Compliance posture. Quake AI holds SOC 2 Type I and Type II attestations and SOC 3. See Compliance and certifications for the platform scope.