Title: WP-AutoInsight
Author: phalkmin
Published: <strong>9 de junio de 2023</strong>
Last modified: 12 de octubre de 2026

---

Buscar plugins

![](https://ps.w.org/automated-blog-content-creator/assets/banner-772x250.jpg?rev
=3509651)

![](https://ps.w.org/automated-blog-content-creator/assets/icon-256x256.png?rev=
3509651)

# WP-AutoInsight

 Por [phalkmin](https://profiles.wordpress.org/phalkmin/)

[Descargar](https://downloads.wordpress.org/plugin/automated-blog-content-creator.4.5.2.zip)

 * [Detalles](https://es.wordpress.org/plugins/automated-blog-content-creator/#description)
 * [Valoraciones](https://es.wordpress.org/plugins/automated-blog-content-creator/#reviews)
 *  [Instalación](https://es.wordpress.org/plugins/automated-blog-content-creator/#installation)
 * [Desarrollo](https://es.wordpress.org/plugins/automated-blog-content-creator/#developers)

 [Soporte](https://wordpress.org/support/plugin/automated-blog-content-creator/)

## Descripción

WP-AutoInsight brings AI content generation into your WordPress dashboard without
a platform subscription attached. It isn’t a SaaS or another subscription service.
You pay for what you use. Connect your OpenAI, Anthropic, Google, or Perplexity 
accounts to the plugin, and your site will generate content at low cost, for a fraction
of what most SaaS tools charge $50-100 per month.

Whether you’re a small business keeping a blog active, an agency managing content
for clients, or a blogger who’d rather talk through ideas than type them, WP-AutoInsight
creates, you review, and _you_ publish.

#### Key Features

 * **Generate content in more ways than you’d expect**
    - Write full blog posts from a keyword list, automatically or on demand
    - Build a Topic Library: reusable topics that generate posts on their own schedule,
      from hourly to weekly, each with its own prompt and model
    - Turn voice notes or meeting recordings into draft posts. Upload audio, get
      a structured article
    - Create infographics from any existing post, saved directly to your Media Library
    - Pull research-backed content through Perplexity Sonar, complete with clickable
      source citations
 * **Choose the AI. Pay the AI directly.**
    - Supports OpenAI, Anthropic Claude, Google Gemini, and Perplexity models. Switch
      models anytime you want
    - Each model shows an estimated cost per post before you choose it
    - Pick a fallback provider: if your primary is rate-limited or down, the post
      finishes on the second one
    - Give any keyword group its own model or length; everything else inherits your
      defaults
    - Your API keys, their actual rates. No markup, no lock-in
 * **Nothing publishes without your approval**
    - Content saves as a draft by default. Review before anything goes live
    - Content History tracks every generated post: which model, which status, when
    - Set tone, keywords, categories, and length once. The plugin will follow your
      rules
 * **Works with everything already on your site**
    - Native Gutenberg block output. Not an HTML blob in a classic editor
    - Yoast SEO and RankMath: focus keywords, meta descriptions, and social excerpts
      generated automatically
    - Featured images via OpenAI GPT Image, Stability AI, or Gemini image generation,
      in the style you choose
 * **For developers**
    - Store API keys in wp-config.php for maximum security, or use WordPress 7.0’
      s native Connectors API
    - Configurable per post type, clean option names, no proprietary lock-in

### Configuration

#### API Keys

You’ll need at least one of the following API keys:
 * OpenAI API key (for GPT models
and GPT Image) * Claude API key (for Claude 4.5 models) * Gemini API key (for Google’s
AI) * Perplexity API key (for web-grounded content with citations) * Stability AI
key (optional, for alternative image generation)

For enhanced security, add your API keys to wp-config.php:
 `php define('OPENAI_API','
your-key-here'); define('CLAUDE_API', 'your-key-here'); define('GEMINI_API', 'your-
key-here'); define('PERPLEXITY_API', 'your-key-here'); define('STABILITY_API', '
your-key-here');

#### Content Settings

 1. Select your preferred AI model
 2. Set your desired content tone
 3. Configure keywords and categories
 4. Adjust token limits and scheduling
 5. Enable/disable image generation
 6. Set up email notifications

### Support

For support, feature requests, or to contribute to development:
 * Read the [Documentation](https://wpautoinsight.phalkmin.me/)*
Visit the [WordPress support forum](https://wordpress.org/support/plugin/automated-blog-content-creator/)*
Submit issues on [GitHub](https://github.com/phalkmin/wp-autoinsight) * For custom
integrations or consulting: phalkmin@protonmail.com * Support development: [](https://ko-fi.com/U7U1LM8AP)

## Capturas

[⌊Plugin settings page - Configure API key, keywords, and other options.⌉⌊Plugin
settings page - Configure API key, keywords, and other options.⌉[

Plugin settings page – Configure API key, keywords, and other options.

[⌊Example generated blog post using Gutenberg blocks.⌉⌊Example generated blog post
using Gutenberg blocks.⌉[

Example generated blog post using Gutenberg blocks.

## Instalación

 1. Upload `wp-autoinsight` to the `/wp-content/plugins/` directory
 2. Activate the plugin through the ‘Plugins’ menu in WordPress
 3. Navigate to ‘WP-AutoInsight’ in your admin menu
 4. Configure your preferred AI service API keys
 5. Set up your content preferences and posting schedule

## FAQ

### How do I get an OpenAI API key?

To use OpenAI models (GPT-4.1, o4-mini), sign up at OpenAI and get your API key:

1. Go to https://platform.openai.com/api-keys 2. Sign up or log in to your account
3. Click «Create new secret key» 4. Copy and paste the key into WP-AutoInsight’s
Advanced Settings

### How do I get a Claude API key?

To use Claude models (Haiku, Sonnet, Opus), you need an Anthropic API key:
 1. Visit
https://console.anthropic.com/ 2. Create an account or sign in 3. Go to API Keys
section 4. Generate a new API key 5. Add it to WP-AutoInsight’s Advanced Settings

### How do I get a Gemini API key?

To use Google’s Gemini models, get your API key from Google AI Studio:
 1. Go to
https://aistudio.google.com/app/apikey 2. Sign in with your Google account 3. Click«
Create API key» 4. Copy the key to WP-AutoInsight’s Advanced Settings

### How do I get a Stability AI API key?

For image generation fallback with Stability AI:
 1. Visit https://platform.stability.
ai/ 2. Create an account and sign in 3. Go to API Keys in your account settings 
4. Generate a new API key 5. Add it to Advanced Settings for image generation

### How do I select AI models?

WP-AutoInsight 4.0 features a visual model selection interface inside the Connections
tab:
 1. Go to WP-AutoInsight > Connections > API Keys 2. Browse model cards organized
by provider (OpenAI, Claude, Gemini) 3. Click on your preferred model to select 
it 4. Each model shows cost tier (Economy/Standard/Premium) and estimated cost per
post 5. Your selection saves automatically

### How can I customize the generated content?

You have extensive customization options:
 – **Keywords**: Set topics and focus 
areas in Content Settings – **Tone**: Choose from Professional, Casual, Friendly,
or Custom tone – **Categories**: Select WordPress categories for posts – **Token
Limits**: Control content length in Advanced Settings – **SEO**: Enable automatic
SEO metadata generation – **Images**: Toggle featured image generation on/off

### Can I use audio files to create blog posts?

Yes! WP-AutoInsight 3.0 includes audio transcription:
 1. Enable Audio Transcription
in settings 2. Upload an audio file to your Media Library 3. Edit the audio file
and click «Transcribe & Create Post» 4. The AI will convert speech to text and create
a formatted blog post 5. Supports MP3, WAV, M4A, WebM, FLAC formats up to 25MB

### How do I create infographics from my posts?

The infographic feature analyzes your content and creates visuals:
 1. Open any 
existing post for editing 2. Look for the «AI Infographic Tools» meta box 3. Click«
Create Infographic» 4. The AI analyzes your content and generates a visual infographic
5. The image is saved to your Media Library automatically

### Can I rewrite existing posts with AI?

Yes, use the AI rewrite feature:
 1. Edit any existing post 2. Find the «AI Content
Tools» meta box in the sidebar 3. Click «Rewrite with AI» 4. The AI will improve
and restructure your content while maintaining the core message 5. Review and publish
the updated content

### How do I manually create posts?

Multiple ways to create posts manually:
 – **Settings Page**: Click «Create post
manually» in Content Settings – **Post List**: Use the «Create AI Post» button on
post list screens – **Quick Creation**: Generate posts from the main dashboard

### Is it possible to schedule automatic content generation?

Yes, WP-AutoInsight offers flexible automation:
 1. Go to Connections > Scheduling
2. Set «Schedule post creation» to Hourly, Daily, or Weekly 3. Configure your keywords
and preferences 4. Posts will be automatically generated and saved as drafts 5. 
Optional email notifications when new posts are created

### Which post types are supported?

You can configure which post types show AI tools:
 1. Go to Content Settings 2. 
Select from available post types (Posts, Pages, Custom Post Types) 3. AI buttons
and tools will appear for selected post types 4. Default is set to standard WordPress
Posts

### How secure are my API keys?

WP-AutoInsight keeps keys out of the places they usually leak:
 – Store API keys
in wp-config.php (or WordPress Connectors on WordPress 7.x) so they never sit in
the database; the Connections tab shows which source is in use – Keys saved through
the settings screen are stored as plain WordPress options, like most plugins do;
treat database backups accordingly – Keys are redacted from the debug log and never
rendered back into the settings form – Every provider call uses HTTPS

### What’s the difference between the AI models?

Each provider offers different strengths:
 – **OpenAI**: Excellent for creative 
and versatile content – **Claude**: Great for analytical and structured content –**
Gemini**: Strong at factual and research-based content – **Perplexity**: Generates
web-grounded content with real source citations, ideal for research-heavy or news-
adjacent posts – **Cost Tiers**: Economy (fast/cheap), Standard (balanced), Premium(
highest quality)

### How does Perplexity work differently from the other providers?

Perplexity searches the web in real time before generating content, then includes
source citations alongside the text. Instead of generating from training data alone,
it pulls from current sources and references them in the post. You can choose how
citations appear: as inline hyperlinks, a references section at the bottom, or both.
You can also set a recency filter to limit sources to the last day, week, month,
or year. A Perplexity API key with an active paid plan is required.

### Can I use multiple AI services together?

Yes, you can configure multiple API keys and switch between models:
 – Set up keys
for different providers in Advanced Settings – Choose different models for different
types of content – The plugin automatically uses the appropriate service based on
your selection

### Does the plugin work with SEO plugins?

Yes, WP-AutoInsight integrates with popular SEO plugins:
 – **Yoast SEO**: Automatic
meta descriptions, focus keywords, and social previews – **RankMath**: Compatible
with meta field generation – Enable «Generate SEO Metadata» in Content Settings 
for automatic optimization

### What happens if content generation fails?

WP-AutoInsight includes robust error handling:
 – Detailed error messages help identify
issues – Automatic fallbacks between different AI services – Content is saved as
drafts to prevent data loss – Error logging helps with troubleshooting – Email notifications
for scheduled generation failures

### How do I get support?

Multiple support channels available:
 – **Documentation**: https://wpautoinsight.
phalkmin.me/ – **WordPress Forum**: https://wordpress.org/support/plugin/automated-
blog-content-creator/ – **GitHub Issues**: https://github.com/phalkmin/wp-autoinsight–**
Direct Contact**: phalkmin@protonmail.com

## Reseñas

No hay valoraciones para este plugin.

## Colaboradores y desarrolladores

«WP-AutoInsight» es un software de código abierto. Las siguientes personas han colaborado
con este plugin.

Colaboradores

 *   [ phalkmin ](https://profiles.wordpress.org/phalkmin/)

[Traduce «WP-AutoInsight» a tu idioma.](https://translate.wordpress.org/projects/wp-plugins/automated-blog-content-creator)

### ¿Interesado en el desarrollo?

[Revisa el código](https://plugins.trac.wordpress.org/browser/automated-blog-content-creator/),
echa un vistazo al [repositorio SVN](https://plugins.svn.wordpress.org/automated-blog-content-creator/)
o suscríbete al [registro de desarrollo](https://plugins.trac.wordpress.org/log/automated-blog-content-creator/)
por [RSS](https://plugins.trac.wordpress.org/log/automated-blog-content-creator/?limit=100&mode=stop_on_copy&format=rss).

## Registro de cambios

#### 4.5.2

 * Fixed: the setup wizard was stuck on an empty screen for new installs and on «
   Restart setup wizard» (broken since 4.3). The first step shows again, and providers
   with a saved key are recognized on restart.
 * Fixed: admin notices no longer show up as unreadable white text inside the wizard
   header.
 * Fixed: going back to the Topic step after creating a topic no longer leaves Continue
   disabled; restarting the wizard no longer resets your chosen model.
 * Fixed: «Manage at Settings  Connectors» links now open the WordPress 7.1 Connectors
   screen.
 * Fixed: the dashboard Composer could publish a post with Draft selected; its fields
   now show exactly what runs.
 * Fixed: Rewrite with AI no longer saves on the server or produces code fences 
   and model notes; the result lands in the editor for you to review, save or undo.
 * Fixed: provider health now updates from the wizard test, the daily check and 
   real generations; Validate works for wp-config and Connectors keys.
 * Fixed: Gemini 2.5 posts failing on hidden thinking tokens; Perplexity reasoning
   leaking into posts; code samples and lists flattened into paragraphs.
 * Fixed: featured-image alt text without an SEO plugin, visible image failure reasons,
   infographics attached without a loopback request and without replacing the featured
   image.
 * Fixed: topics gain optional focus keywords and local-time dates; auto-pause counts
   real failures; stuck «Queued» jobs are failed with a WP-Cron hint; bulk stops
   after a key error.
 * Fixed: restarting the wizard kept Publish; Reset keeps API keys; many copy, i18n
   and accessibility fixes.
 * Security: job status scoped to the creator; API keys redacted from debug logs.
 * Fixed: regressions found in the 2026-10-10 hunt, before the tag: bulk SEO jobs
   wrongly failed as stale, a false «WP-Cron has not fired» dashboard warning, three
   content-converter cases that lost text, health writes from typed keys, and a 
   dozen smaller fixes.

#### 4.5.1

 * Fixed: turning AI off in WordPress (7.0+) now also stops scheduled topics and
   queued jobs; they resume when AI is back on.
 * Fixed: after an automatic update, scheduled posts no longer run with new defaults
   before an administrator visits the dashboard.
 * Fixed: on sites using a server cron job, each scheduled post now runs as its 
   own background job.
 * Security: scheduled posts become drafts when their author can no longer publish;
   audio tools check that you can edit the selected file.

#### 4.5.0

 * Provider fallback: pick a second provider under Connections  API Keys. When your
   primary hits a rate limit, an outage, or a network error, the post is retried
   once on the fallback and the log says which provider finished it. Invalid keys
   never fall back.
 * Per-keyword-group model and length: an «Advanced» block on each group lets you
   choose a different model or length for that group; blank inherits your global
   setting.
 * Image style: choose a look for featured images (editorial photography, flat illustration,
   3D render, minimalist line art, watercolor, cinematic, or your own description).
   Prompts now also ask for no text, watermarks, or logos.
 * «Remove saved key» beside each stored API key, with confirmation.
 * Failed jobs show a one-line «what to do next»; Bulk Generate ends with a created/
   failed summary and a log link; Topics «Run now» links to the log.
 * Auto-saving fields are labelled «Saves automatically»; the wp-config key tip 
   is tucked behind «Advanced»; support links moved to Settings  Advanced.
 * Fixed: the «Post from Audio» tile now opens the post editor where the upload 
   lives; unreachable wizard steps removed; provider logos labelled for screen readers.
 * For developers: extension hooks (`abcc_before_generate`, `abcc_after_generate`,`
   abcc_post_inserted`, `abcc_generation_prompt`, `abcc_generation_result`, `abcc_settings_schema`,`
   abcc_admin_tabs`) — public documentation follows in 4.6. The `abcc_post_inserted`
   context now also carries `served_provider` / `served_model`.
 * Security: users whose role cannot publish a post type now get drafts from the
   Composer, matching the editor; generation jobs pin their draft/publish decision
   when queued, so deleting a Topic or changing the default while a job waits cannot
   publish a draft; unchecking Editors under Permissions now actually revokes their
   access.
 * Fixed: «Transcribe Only» no longer creates (and bills) a post; «Regenerate as
   New Draft» always produces a draft; the «Remove saved key», «Restart setup wizard»,«
   Reset All Settings» and «Delete All Generation History» buttons now reliably 
   submit their action (the form script was disabling the clicked button before 
   the browser read it); adding a custom template no longer saves it empty; removing
   then adding a keyword group no longer overwrites a surviving group; keywords 
   or tones with quotes or accents no longer break «Regenerate»; when a fallback
   provider writes the body, Perplexity citations are linked and the post records
   the model that actually wrote it.

#### 4.4.0

 * Content language: posts, titles, SEO metadata, and audio intros are written in
   your site’s language automatically. Override it under Content  Keywords  Writing
   Style. Old custom templates without a {language} placeholder still come out right.
 * Truncation repair: posts that hit the model’s token limit are trimmed to the 
   last complete section and given a proper closing paragraph instead of stopping
   mid-sentence. The generation log notes when this happened.
 * Generation errors now say what went wrong and what to do — a rejected key, a 
   rate limit, an oversized prompt, and a provider outage each get their own actionable
   message.
 * Featured image failures are visible: the post edit screen shows why the image
   failed, with a Retry button.
 * New «Restart setup wizard» button in Settings  Advanced (keeps your settings 
   and API keys).
 * Saved API keys are no longer displayed in settings pages; leave the field blank
   to keep your existing key.
 * Security: onboarding connection test requires administrator capability; rewrite/
   regenerate check per-post edit permission; Gemini keys moved from the request
   URL to a header; admin status areas no longer inject server-supplied HTML.
 * Fixed: featured images failing to attach on local and firewalled sites; Perplexity
   citations attaching to the wrong post under concurrent schedules; prompts being
   HTML-mangled before sending; bulk SEO firing one request per post; cron schedules
   queried on every page load; SEO parsing failing on braces in prose; the dashboard
   SEO Refresh tile linking nowhere; the wizard redirecting away from its completion
   screen; hover-only error messages; misleading «Settings saved» notices after 
   destructive actions; endless job polling; the settings JSON export producing 
   a broken page instead of a download; Enter breaking multi-line wizard fields.
 * Post length now reads in approximate words; model names show in readable form;
   tooltips work with keyboard focus; confirm dialogs are translatable.

#### 4.3.0

 * Dashboard reframed into a Command Center: the new «Composer» card shows exactly
   what your next post will use — source, template, model, and draft/publish — and
   lets you change any of it in place before generating. Per-post choices stick 
   as the new default without touching your global settings.
 * New capability grid on the dashboard surfaces everything the plugin can do (Topics,
   Bulk Generate, Post from Audio, SEO Refresh, Featured Images, Infographics, Schedule,
   Providers) at a glance.
 * Post from Audio: upload a recording and choose how it becomes a post — keep your
   words with an AI-written intro and title, or have AI rewrite the recording into
   a full article. Output honors your draft/publish default; if a rewrite fails,
   your transcription is never lost.
 * Refresh: new bulk «Regenerate SEO» action on the Posts list — select posts and
   queue background jobs to regenerate titles, meta descriptions, and focus keywords
   with your chosen model. Works with Yoast and Rank Math.
 * Rewrote first-run onboarding into a focused flow: connect a provider, choose 
   your draft/publish default, optionally create your first Topic, and discover 
   Post from Audio. «Skip to settings» is available on every step.
 * Updated the AI model lineup: Claude Opus 4.8, OpenAI GPT-5.4 / GPT-5.4 mini /
   GPT-5.5, Gemini 3.5 Flash, and Perplexity Sonar Deep Research. GPT-4.1 models
   remain available, now labeled «(legacy)». New installs default to GPT-5.4 mini.
   Models with highly variable cost show a warning in the selector.
 * Fixed: Post from Audio now uses your selected text model’s API key (it previously
   reused the OpenAI transcription key, silently failing for Claude/Gemini/Perplexity
   users), embeds the audio player, splits the transcript into paragraphs, and cleans
   Markdown out of the title.
 * Fixed: onboarding now saves your Publish/Draft choice correctly and adds a «Skip
   to settings» link to the first step.
 * Fixed: Bulk SEO Regeneration uses each job’s chosen model key and skips posts
   you can’t edit.
 * Fixed: GPT-5.x models failed with «Unsupported parameter: ‘max_tokens’» — the
   OpenAI client now sends the parameters these models require and omits the ones
   they reject.
 * Fixed: GPT-5.x content generation returned empty results (a silent «Content generation
   failed») because the models spent the whole token budget on internal reasoning.
   Reasoning is now disabled for GPT-5.x requests so every token goes to your post,
   and empty completions from any provider are logged with details instead of failing
   silently.
 * Fixed: posts came out as one long paragraph when the model returned the article
   on a single line; content is now split into proper heading/paragraph blocks for
   all providers.
 * Fixed: post titles could end up as the model’s preamble («Here are some catchy
   blog post titles…») when the AI replied with a list of title options (seen with
   Gemini). The plugin now asks for exactly one title and picks a clean title out
   of list-style replies.
 * Fixed: the WordPress admin footer overlapped the dashboard’s «At a glance» section
   due to a stray closing tag in the dashboard template.
 * Fixed: new models (GPT-5.x, Gemini 3.5 Flash, Claude Opus 4.8, Sonar Deep Research)
   were silently limited to a 4,096-token budget, producing much shorter posts than
   configured.
 * Fixed: a PHP deprecation notice from the job log when a generated post had been
   deleted.

#### 4.2.0

 * New Topic Library tab: reusable topics with per-topic schedule (hourly to weekly),
   draft/publish override, and model override. Pause, resume, edit, or run any topic
   on demand; topics that fail repeatedly pause themselves and tell you why.
 * Generated posts now save as drafts by default, with a new Draft/Publish setting.
   Existing installs keep their current behavior automatically and see a one-time
   notice.
 * Fixed: duplicate posts from one generation job, provider health false «connected»
   status, Claude API error handling, stray `<title>`/`[SEO]` lines in content, 
   Gemini image API key encoding.
 * Fixed: OpenAI image generation failing after DALL-E 3’s deprecation. Now uses
   the GPT Image models (gpt-image-1, Mini, 1.5), selectable on the Images tab; 
   existing installs recover automatically.
 * Internal: unified API call layer for all text providers with consistent error
   handling and truncation detection; onboarding wizard split into step partials.

#### 4.1.1

 * Keyword Groups now pick **one** keyword at random per generated post instead 
   of mashing every keyword into a single article. A «Chemistry» group with ten 
   keywords becomes a queue of ten focused topics rather than one over-broad post.
 * New `{keyword}` template placeholder holds the keyword chosen for the current
   post; `{keywords}` still renders the full comma-separated group for templates
   that want it. The default template now uses `{keyword}` so existing setups get
   the focused behavior automatically.
 * Generation tracking meta records which keyword was selected; the post-editor 
   meta box now shows the focus keyword alongside the source group.
 * Fixed: the Default Template now displays its prompt body in a read-only textarea
   on the Content settings page instead of an empty placeholder.

#### 4.1.0

 * WP 7.0 Connectors API integration now works correctly (previously broken hook
   and registry calls removed)
 * Admin notice shown when WordPress has AI disabled at site level (WP 7.0)
 * Claude models refreshed: Sonnet 4.6 and Opus 4.7 (1M token context window each)
 * Connections tab shows read-only legacy key reference when WP Connector is active
   for a provider
 * Contextual «Learn more» links added to API Keys, Scheduling, and Bulk Generate
   sections
 * ~1.9 MB runtime reduction: Gemini SDK replaced with direct wp_remote_post() calls,
   Composer vendor removed from runtime

#### 4.0.1

 * Fixes to «Validate» button for API keys
 * Adding documentation URL to the dashboard

#### 4.0.0

The biggest settings redesign since launch. Everything is easier to find, the dashboard
tells you what’s happening without digging, and settings that used to feel buried
now live exactly where you’d look for them.

**A real dashboard, finally.**
 The plugin now opens to a dashboard instead of a
blank settings form. You can see your scheduling status at a glance, check the last
few generation jobs without leaving the page, and get to common actions in one click.
Provider health is summarized right there — green/yellow/red, with timestamps so
you know if a key was last verified today or a week ago.

**Settings that make sense.**
 The old flat tab layout is gone. Settings are now
split into five focused sections: Content (keywords, templates, writing style), 
Media (images, audio, infographics), Connections (API keys and scheduling), Settings(
post types, permissions, advanced), and the Dashboard. Everything that previously
required hunting through a long page now has its own tab.

**Changes that save themselves.**
 Toggles and sliders across the settings pages
now save automatically as you adjust them, with a small «Saved» confirmation. No
more scrolling to the bottom to hit a save button after every change. Fields that
affect scheduling still require an explicit save (and warn you if you leave without
saving).

**Per-role access control.**
 You can now control which WordPress roles can access
AI generation tools. Authors, editors, or just administrators — your call. Administrators
are always enabled and can’t be unchecked.

**Provider health checks run daily.**
 The plugin now runs a background health check
on your connected providers once a day and caches the result. The dashboard shows
the freshness of each check so you know whether a «connected» status is from this
morning or three days ago.

**Debug mode and settings export.**
 The Advanced tab now has a one-click debug 
logging toggle (no more editing wp-config.php for basic troubleshooting) and a JSON
export button so you can back up or migrate your settings.

For the full changelog of earlier versions, see CHANGELOG.txt.

## Meta

 *  Versión **4.5.2**
 *  Última actualización **hace 4 días**
 *  Instalaciones activas **10+**
 *  Versión de WordPress ** 6.8 o superior **
 *  Probado hasta **7.1.3**
 *  Versión de PHP ** 7.4 o superior **
 *  Idioma
 * [English (US)](https://wordpress.org/plugins/automated-blog-content-creator/)
 * Etiquetas:
 * [ai content](https://es.wordpress.org/plugins/tags/ai-content/)[anthropic](https://es.wordpress.org/plugins/tags/anthropic/)
   [google-ai](https://es.wordpress.org/plugins/tags/google-ai/)[openai](https://es.wordpress.org/plugins/tags/openai/)
   [Perplexity](https://es.wordpress.org/plugins/tags/perplexity/)
 *  [Vista avanzada](https://es.wordpress.org/plugins/automated-blog-content-creator/advanced/)

## Valoraciones

Aún no se han enviado valoraciones.

[Tu valoración](https://wordpress.org/support/plugin/automated-blog-content-creator/reviews/#new-post)

[Ver todas las reseñas](https://wordpress.org/support/plugin/automated-blog-content-creator/reviews/)

## Colaboradores

 *   [ phalkmin ](https://profiles.wordpress.org/phalkmin/)

## Soporte

¿Tienes algo que decir? ¿Necesitas ayuda?

 [Ver el foro de soporte](https://wordpress.org/support/plugin/automated-blog-content-creator/)