1. Added support for X.AI API - the grok-beta model is available (with streaming and tool call support)
2. Added support for Nvidia AI language models
3. New DIY OmniBlock type - add your own custom PHP code in a WordPress filter and execute it in the OmniBlock sequence, getting results and parsing (filter name template: aiomatic_diy_omniblock_OMNIBLOCKID)
4. Added more OmniBlock default templates
5. WordPress 6.7 compatibility
6. Added the ability to filter OmniBlock templates listing by category
7. Added a new OmniBlock type: Plain Text OmniBlock, which can hold textual values or be used to process shortcodes and Spintax (supports nested shortcodes also, to import content generated by other of CodeRevolution's plugins)
8. Allow assignation of custom namespaces to created embeddings
9. Allow the usage of different namespaces for embeddings in different parts of the plugin (to add different sets of extra knowledge to content creators, different chatbot, content editors, etc)
10. User aware embeddings - you can use additional shortcodes in embeddings, to get data specific to current user
11. Added the ability to automatically index embeddings from newly published post comments/product reviews
12. AI Forms new input field type: file upload - allow users to upload files and the AI will be able to analyze them and parse their content. Supported file types: all textual file types, including txt, json and pdf
13. AI Forms more new input field types added: color, range, date, time, datetime, month, week selectors
14. User aware and content aware AI Forms - you can use additional shortcodes in AI Forms prompts, to get data specific to current user or to currently viewed post
15. Added the new Anthropic Claude model: claude-3-5-haiku-20241022
16. Added a Prompt Library to the AI Playground, it contains a large list of useful prompts you can try
17. Added a Model Comparison Tool to the AI Playground, allowing you to test and compare any AI model
18. Royalty free image attribution text now can be used in content creation prompts
19. Added support for Bing SERP API - for AI Internet Results and Related Searches
20. Added support for Serper.dev API - for AI Internet Results and Related Searches
21. Added support for SpaceSerp API - for AI Internet Results
22. Added support for Google Programmable Search Engine API for Royalty Free Images search
23. Added support for tools/function calling for Groq AI models
24. D-ID talking avatar now loads much faster and also in a much more reliable way
25. Removed "The Best Spinner" from supported article spinners list, this service is not available any more
1. Added AI content writer integration for Elementor Text Editor and Headlines widgets
2. Fixed Azure Assistants not working if some special model deployment names were used
3. Added the ability to set the default AI model which will be used for requests
4. Sorted OpenRouter model list alphabetically
5. Various fixes and improvements
1. Ollama embeddings now work with more models
2. Fixed Anthropic Claude streaming
3. Added new Anthropic Claude Sonnet model: claude-3-5-sonnet-20241022
4. Added the ability to hide Aiomatic's features on the single page/post editor
5. On deactivation, the plugin will prompt about purchase code deactivation (revoking)
6. On deactivation, the plugin will be able to remove also its internal data (settings)
7. More parameters can be passed to Replicate Flux API
8. Fixed the [aicontent] shortcode renderer mechanism, it will no longer render other shortcodes on the page if the [aicontent] shortcode is not found in the content
9. Fixed error in AI Forms when combined with some browsers
10. Updated translations
1. Added support for Azure OpenAI Assistants API
2. Added support for Ollama embedding models, check available models, here: https://ollama.com/search?c=embedding
3. Added AI Vision (image recognition) support for Ollama, you can check AI models supporting vision, here: https://ollama.com/search?c=vision
4. Added Tool Calling support for Ollama, you can check AI models supporting tool calls (function calls), here: https://ollama.com/search?c=tools
5. Added new Google AI model: gemini-1.5-flash-8b-latest
6. Added support for Google Gemini AI models Vision feature
7. Added support for Anthropic Claude AI models Vision feature
8. Added support for Groq AI models Vision feature
9. Added support for OpenRouter AI models Vision feature
10. Added support for Azure OpenAI models Vision feature
11. Added Groq new 'preview' models (including llama 3.2 family models)
12. Added support for o1 family models in Azure OpenAI API
13. Added the ability to select the used Azure OpenAI API version
14. Bulk AI Post Creator - YouTube to Posts module, allow to get videos from YouTube playlists
15. Fixed chatbot text-to-speech and text-to-video, no longer including also HTML tags, only plain text
1. Added the ability to store AI prompts on OpenAI's part for different modules of the plugin (chatbot, forms, AI Content Editor, Bulk AI Post Creators or Globally). This stored data can be used for custom model training, model distillation or evaluations
2. Automatic comment/review writer improvements in user name generator and comment source IP address
3. Added more API options for Replicate AI models
4. Added the ability to show a real time waveform animation, each time the chatbot speaks
5. Added support for the latest omni-moderation-latest text moderation model
6. Fixed rare fatal issue at first plugin install
7. Fixed an incompatibility with the WoodMart theme
1. Added support for Replicate Flux AI image models. Go and play with Flux, it's really nice! (other AI image models available: bytedance, stabilityai, black forest labs, ai forever, datacte, fofr, tstramer, playgroundai, lucataco, batouresearch, adirik, nightmareai)
2. Added support for AI Forms visual themes
3. [aiomatic-text-editing-form] improvements, now it can get a predefined prompt, which should be used to edit the user input. Also, more parameters were added to this shortcode: prompt, edit_placeholder, instruction_placeholder, result_placeholder, submit_text, enable_copy, enable_speech
4. Fixed internal links not appearing in some cases
5. Added the ability to 'Create Post' OmniBlock to skip posts which are already published with the same title
1. Added new OpenAI reasoning models: o1-preview, o1-preview-2024-09-12, o1-mini and o1-mini-2024-09-12 - give them a try, they are useful in coding and other technical queries (but still in beta)
2. Added new Default Chatbot Personas
3. Added support for Azure Text-to-Video Streaming (as a D-ID alternative)
4. Added "Internet Search" OmniBlock
5. Improved [aicontent] shortcode, now it is able to process other shortcodes inside of it
6. Added the ability to the [aicontent] shortcode to use the new repeat_for_each_line parameter, to repeat the prompt sent to the shortcode for each line of this parameter. You can get the current line from inside the prompt, using the %%current_line%% shortcode
7. Removed deprecated AI models: gpt-3.5-turbo-0301, gpt-3.5-turbo-0613 and gpt-3.5-turbo-16k-0613
8. AI Content Editor - added the ability to filter automatically edited posts by author user ID
9. Fixed chatbot auto scroll to bottom on new message not working any more on new Chrome versions
10. Improved Bulk AI Post Creator image prompt processing when AI generated images are used
11. Removed unwanted ```html from AI generated texts
12. Fixed blank titles being created on some rare servers in Bulk AI Post Creators and OmniBlocks
13. Fixed logging and statistics error
1. Improved API Key rotation in case a key is rate limited for all AI API service providers
2. Added support for the chatgpt-4o-latest model from OpenAI
3. Added support for AI image creation (OpenAI, Stable Diffusion and Midjourney) and royalty free image creation, to the [aicontent] shortcode
4. Added support for the new llama-3.1-sonar-huge-128k-online model in Perplexity AI
5. Fixed statistics issues appearing in some rare cases
6. Fixed Media Library extension issues when the SSL certificate was not valid on the server
7. Fixed Stable Diffusion image editing in case of some image types
8. Fixed SQL injection vulnerabilities
1. Added support for AI Content Editor templates - allowing quick selection of multiple ways how to edit different posts
2. Added the ability to define multiple automatic AI Content Editors, editing different content (post types, categories, tags), in different ways (using different prompts or different operations)
3. Added support for the new gpt-4o-2024-08-06 model
4. Added support for training the gpt-4o-mini model
5. Improvements to user interface and admin menu
6. Code refactoring, for consistency
7. Updated translations
1. AI Content Editor: added the ability to transform audio files to text using OpenAI's Whisper model
2. Added support for Groq AI
3. Added support for base64 encoded pdf files
4. Added for the topic based Bulk AI Post Creator the ability to import data for post topics from RSS feeds and txt documents
5. Updated Perplexity AI model list, added new Llama 3.1 models