Description

VozCaster is a Telegram bot that turns voice notes into fully published podcast episodes on your WordPress site. This plugin — Connector for VozCaster — is the WordPress side of the system: it receives episodes from the bot and publishes them using PowerPress.

You send a voice note to the bot, and a published draft (or live episode) appears on your site with audio, title, content and featured image. No editing, no manual file uploads, no opening the WordPress dashboard.

What it does

Receives audio from the VozCaster bot and uploads it to your WordPress media library.
Creates podcast episodes using PowerPress, with episode and season numbering managed automatically.
Supports multiple podcasts on the same site — works with any PowerPress custom channels you have configured.
Per-podcast settings: title prefix, intro/outro audio, post footer (signature), category mapping.
Granular permissions: choose which WordPress users can publish to which podcast.
Token-based authentication so the bot never sees your WordPress password.

How it works

Install and activate this plugin on your WordPress site.
In Settings Connector for VozCaster, authorise one or more WordPress users to publish via the bot.
Open Telegram, talk to the VozCaster bot, run /conectar and link your WordPress site.
Send a voice note or audio file to the bot. The episode appears on your site, ready to review or publish.

Requirements

WordPress with the PowerPress plugin installed and activated.
A Telegram account.
A WordPress user authorised to publish via the bot (configurable in the plugin settings — any role works as long as the administrator marks the user as authorised).

External Services

This plugin requires an external service to function: the VozCaster bot, a Telegram bot operated by the plugin author (Miguel Ángel Terrón Bote). The plugin is the WordPress-side component; the bot performs audio processing and orchestrates publication.

The following data is transmitted:

From the bot to your WordPress site, via this plugin’s REST API
- Audio files received by the bot (sent by you via Telegram).
- Optional featured images.
- Episode metadata: title, description, content, episode number, season, feed slug.
- An authentication token issued to your WordPress user during the connection flow.
These calls are initiated by the bot, not by the plugin. They reach your site through standard REST endpoints registered by this plugin under /wp-json/vozpress/v1/.
From your WordPress site to the bot (only during the user authorisation flow)
- When a WordPress user connects via /conectar in Telegram, your browser is redirected to your WordPress login. After login, this plugin issues a token and returns it to the bot to complete the handshake.
From the bot’s server to Telegram’s servers
- User-sent voice messages and audio files pass through Telegram’s infrastructure as part of normal bot communication. See Telegram’s Privacy Policy: https://telegram.org/privacy
From the bot’s server to AI providers (text only)
- To draft the episode title, description and content, the bot sends the transcribed text of your recording to Anthropic (Claude).
- If you request a cover image, the bot sends a text prompt to OpenAI to generate the image.
- Only text is sent, and only when content or image generation is requested. Your audio file is never transmitted to these providers.

Audio is transcribed locally with a Whisper model on the bot’s server; other audio processing (DeepFilter noise reduction, ffmpeg, and an Ollama model used only as a local fallback) also runs locally. The audio file itself is not sent to any third-party AI service.

Service URLs and policies:

VozCaster (bot service): https://vozcaster.com
VozCaster privacy policy: https://vozcaster.com/privacy
VozCaster terms of service: https://vozcaster.com/terms
Telegram Messenger: https://telegram.org
Telegram privacy policy: https://telegram.org/privacy
Telegram terms of service: https://telegram.org/tos
Anthropic (Claude, AI text generation): https://www.anthropic.com
Anthropic privacy policy: https://www.anthropic.com/legal/privacy
Anthropic terms of service: https://www.anthropic.com/legal/consumer-terms
OpenAI (AI image generation): https://openai.com
OpenAI privacy policy: https://openai.com/policies/privacy-policy
OpenAI terms of service: https://openai.com/policies/terms-of-use

Screenshots

Podcast access permissions and the list of WordPress users authorised to publish from the bot.

Per-podcast settings: title prefix and season numbering, intro/outro audio, audio mix levels and the post footer. — Podcast access permissions and the list of WordPress users authorised to publish from the bot.

Installation

Upload the connector-for-vozcaster folder to /wp-content/plugins/, or install via Plugins Add New Upload Plugin.
Activate the plugin through the Plugins menu in WordPress.
Go to Settings Connector for VozCaster and authorise the WordPress users that will publish via the bot.
Open Telegram, search for the VozCaster bot, send /conectar and follow the instructions to link your site.

FAQ

Does it work with PowerPress?

Yes. PowerPress is a hard requirement — the plugin uses PowerPress to register episode metadata, enclosures, season and episode numbers.

Do I need a paid account to use it?

This plugin is free and open source. The VozCaster bot it connects to offers a free tier and a paid (Pro) tier. Basic episode publishing works on the free tier; advanced features — automatic transcription, AI-assisted content generation and audio noise reduction — require a Pro subscription. See https://vozcaster.com for the current plans.

Where is my audio stored?

Permanently, only in your own WordPress media library. The bot processes the audio on a private server operated by the author (located in Spain) and deletes the temporary files after publishing.

Does the bot send my audio to third-party AI services?

Your audio is never sent to a third-party AI service. Transcription runs on a local Whisper model on the bot’s server, so the audio file itself stays on that server (and on Telegram, which carries your message).

To draft the episode title, description and content, the bot sends the transcribed text to Anthropic’s Claude API. If you ask the bot to generate a cover image, the image prompt is sent to OpenAI’s image API. These providers receive text only, and only when you request content or image generation. See the “External Services” section below for the providers and their privacy policies.

Can I connect more than one WordPress site to the bot?

Yes. Each site needs its own installation of this plugin and its own connection from Telegram (/conectar for each site).

Can a single WordPress site host more than one podcast?

Yes. The plugin works with PowerPress’s multi-podcast (custom channels) configuration. From Telegram you switch between podcasts with /feed.

Who can publish from the bot?

Any WordPress user whose ID is in the authorised users list (configured in Settings Connector for VozCaster). Role does not matter — what matters is whether the administrator has authorised them.

How do I get support?

For documentation, see the VozCaster manual. For issues or questions, email info@vozcaster.com.

Reviews

There are no reviews for this plugin.

Contributors & Developers

“Connector for VozCaster” is open source software. The following people have contributed to this plugin.

Miguel Angel Terrón

“Connector for VozCaster” has been translated into 1 locale. Thank you to the translators for their contributions.

Translate “Connector for VozCaster” into your language.

Interested in development?

Browse the code, check out the SVN repository, or subscribe to the development log by RSS.

Changelog

1.5.14

Fixed a fatal error on audio episode upload: wp-admin/includes/media.php (which defines wp_read_audio_metadata()) was not loaded in the REST API request context, causing attachment metadata generation to fail for audio files.

1.5.13

Added plugin banner, icon and screenshots for the WordPress.org listing.

1.5.12

Documentation: corrected the “paid account” FAQ to describe the free and Pro tiers of the VozCaster bot (removed outdated “beta” wording).

1.5.11

Settings-changing REST endpoints (global intro/outro, mix settings, plugin options, season) now require an administrator capability, not just an authorised bot user.
The media picker script on the settings screen is now enqueued via admin_enqueue_scripts and wp_localize_script instead of inline output.
Core admin includes in the REST upload helpers are loaded only where needed, immediately before the function that uses them.

1.5.10

Passes Plugin Check with no errors. File operations during chunked audio assembly now use the WordPress Filesystem API; output escaping, input unslashing and wp_parse_url() applied where flagged.
The Telegram connection result pages are now fully translatable (English source strings with an updated Spanish es_ES translation).

1.5.9

Hardened the REST authentication token handling (sanitised header input).
Documentation: the “External Services” section and FAQ now disclose the AI providers used by the VozCaster bot to generate episode text (Anthropic) and cover images (OpenAI). Your audio is never sent to these providers; only text is.

1.5.8

Spanish (es_ES) translation added. The plugin automatically displays in Spanish on WordPress installations set to Spanish.

1.5.7

Renamed from VozPress Connector to Connector for VozCaster to align with WordPress.org plugin guidelines on naming.
Endpoint /episode/next-number now also returns the chapter number within the current season so the bot can name uploaded media files using the season+chapter scheme.

1.5.6

/episode/next-number includes episode_no_in_season for season-aware media file naming.

1.3.2

Feed detection now includes PowerPress “category podcasting” feeds. Each linked category is exposed as a feed using the category slug.
assign_podcast_category() no longer creates new categories — prevents duplicate “Podcast” categories on sites with custom slugs.

1.3.1

Season and chapter-within-season numbering derived from the previous published episode’s PowerPress metadata, instead of the previous year-based heuristic.
Multi-feed detection now reads from powerpress_general.custom_feeds — the per-podcast permissions UI in wp-admin now appears correctly on sites with PowerPress custom channels.

1.3.0

All admin UI strings translated and properly wrapped in i18n functions.
Added == External Services == section to readme.

1.2.x

Multi-feed support: per-feed permissions, automatic category assignment, per-feed episode counters.
Chunked audio upload with gzip compression for large files (>25 MB).
Manual season increment via REST endpoint and /temporada command in the bot.

1.1.x

Intro and outro mixing with configurable ducking.
Settings endpoints for remote configuration from the bot.
Transcript attachment via REST endpoint.

1.0.x

Initial release: /ping, /auth/verify, /media/audio, /media/image, /episode REST endpoints.
Token-based authentication using wp_check_password().
Settings page with access token and episode log.

Description

What it does

How it works

Requirements

External Services

Screenshots

Installation

FAQ

Does it work with PowerPress?

Do I need a paid account to use it?

Where is my audio stored?

Does the bot send my audio to third-party AI services?

Can I connect more than one WordPress site to the bot?

Can a single WordPress site host more than one podcast?

Who can publish from the bot?

How do I get support?