Limited-Time Offer

Get Extra 20% off on FullCharge Yearly Plan — available for first 10 users only. Don’t miss out! 🚀

TrackWP

MxChat · Version history

MxChat changelog and version history

Latest version
3.2.28
Nov 6, 2026 (2 days ago)
Releases
158
since 2024
In the last year
61
about one every 6 days
Tested up to
WordPress 7.1.4
PHP 7.4 or later

Releases each year

  • 202436 releases
  • 202572 releases
  • 202650 releases

202650 releases

  1. MxChat 3.2.28Latest

    • New: Developers can replace the error messages visitors see in the chat with their own wording using the new mxchat_error_response filter, and answer knowledge base queries from their own vector store using the new mxchat_pre_vector_query filter.
    • Fixed: With Contextual Awareness on, security firewalls such as Cloudflare or Wordfence no longer block chat messages. The page text is sent in a firewall-safe form without hidden page content, and a blocked message is retried once without it.
    • Improved: When a security firewall still blocks a message, the chat says so instead of reporting an expired session.
    • Fixed: The chatbot again remembers earlier messages that contain an ampersand, bold text, links or code, instead of leaving them out of the conversation and answering earlier questions again.
    • Fixed: With Pinecone, the Select WordPress Content list now looks up only the posts on the page you are viewing, so it opens far faster on a large knowledge base and shows each post’s full chunk count. It shows the data of the bot you selected, and a post imported before a move to https or a change…
    • Fixed: The In Knowledge Base filter no longer lists every post when nothing has been imported.
    • Fixed: With Citation Links on, answers drawn from an uploaded document or PDF with no Source URL now name the file, and the page of a PDF, in plain text. Before, the file was named only when a visitor asked where the answer came from.
    • Fixed: On older databases, the tables for transcripts, ratings, translations, Pinecone roles and the import queue could be refused with no message, so those features quietly did nothing. They are now created, a table that could not be created is tried again instead of being recorded as done, and a…
    • Fixed: The Test Streaming Compatibility button no longer adds a conversation to Transcripts.
    • Fixed: Provider outages are handled more cleanly. A request that timed out is no longer repeated, so a stalled provider holds a chat for about two minutes instead of six. Claude and Gemini errors are no longer saved as bot replies in Transcripts, visitors no longer see technical details when a cust…
    • Fixed: On a site that reaches the internet through a proxy set in wp-config.php, streamed replies, the Editor Assistant and article images now use that proxy instead of trying to connect directly.
    • Fixed: Behind Cloudflare, visitors no longer share one limit when the chat refreshes its security check, and click tracking and the new-chat email record the visitor’s address instead of Cloudflare’s.
    • Fixed: On narrow screens an MxChat information notice no longer covers the open admin menu and blocks its links. Notices now sit in the page’s content column on the Theme and Admin AI Chat screens too.

  2. MxChat 3.2.27

    • Fixed: Security hardening and general security improvements.
    • New: Delete Leads After, on the Transcripts Notifications screen, deletes a lead and its conversations after a period without chats that you choose (default Never). Deleting a lead or chat also removes its ratings and link clicks.
    • New: A Generate Image (xAI) action and AI tool create images in the chat with xAI Grok Imagine, beside the OpenAI and Gemini ones.
    • New: Developer filters: mxchat_filter_roles_rate_limit picks the role a multi-role user is rate-limited by, mxchat_stream_request can change a streamed chat request, and mxchat_lead_retention_days adjusts lead auto-delete.
    • New: The REST API transcripts endpoint finds answers given without knowledge with knowledge_used=no, each row reports sources_used and knowledge_used, and filtering by role assistant now returns the chatbot’s replies.
    • Fixed: With the Complianz integration on, the chat never opens without cookie consent. Chats placed in a page now wait for consent like the floating chat, a delayed Complianz script is waited for, and withdrawing consent closes the chat.
    • Fixed: Transcripts, the reloaded chat and the transcript email now show each reply as the visitor saw it, with product cards, forms and images in one message, the pre-chat name and email, and no stray code or empty messages.
    • Fixed: Knowledge base sync now follows synced patterns, posts that are unpublished, trashed or restored, and PDFs in ACF Blocks, and pages built with Breakdance, Oxygen, Bricks or Divi 5 show their published text in the import list.
    • Fixed: Saving settings no longer swaps an unrecognised chat model or an older embedding model, retired chat and image models use their replacement, and o-series models, Custom provider streaming and web search queries work on newer models.
    • Fixed: With delayed script loading the chat opens on the first click, the first open no longer hits a Too Many Requests error, replies no longer leave extra blank lines, and an add-to-cart critical error is gone.
    • Fixed: Many layout fixes in the MxChat admin pages on phones, tablets and wide screens, WordPress notices no longer cover the MxChat header and menu, and the Actions page reports embedding errors instead of saving phrases that cannot match.
    • Improved: Lower AI costs and clearer usage: one embedding call per reply instead of two, Transcripts show the tokens spent finding each answer (WooCommerce product searches too), and every model picker uses one list.

  3. MxChat 3.2.26

    • Security: Hardened how chat messages, code, visitor feedback, translations and import details are displayed, how PDF links in the chat are opened, and the permission checks on several admin requests. Update recommended.
    • Security: The pre-chat form and consent are now enforced on the server, and chat limits are applied more strictly. Update recommended.
    • New: GPT-6.1 Sol, Claude Haiku 5.5, Claude Sonnet 5.5, Claude Opus 5.5, Claude Fable 5.1, Grok 4.7, Gemini 4 Argon, DeepSeek Flash and the Gemini Embedding 2 embedding model.
    • New: Each provider’s recommended model is now its fast, affordable choice for customer chat, such as GPT-6.1 Sol, Claude Haiku 5.5 and Gemini 3.8 Flash. Your current model is not changed.
    • New: Models a provider discontinues are replaced automatically, with a notice naming the old and new model, and models with an announced retirement date move on that date. Older models are no longer offered for new picks. Embedding models are never switched for you.
    • New: Transcripts show the model and tokens each reply used, including cached tokens, and OpenAI requests are set up to hit the prompt cache more often.
    • New: Staff can read Transcripts without being administrators through the mxchat_view_transcripts capability.
    • New: The mxchat_should_index_post filter excludes posts from the knowledge base, Auto-Sync can keep only already-imported entries up to date, and add-ons can generate images with mxchat_generate_media_image().
    • Improved: The knowledge base imports ACF Blocks and pages built with Breakdance, Oxygen, Bricks, Divi and WPBakery, and Re-import affected entries rebuilds pages stored without their text and products missing their tab text.
    • Improved: Auto-Sync sends one embedding request per save instead of two or three, and none when nothing changed. Editing a synced pattern updates the pages that use it.
    • Improved: The chat launcher appears straight away on cached pages and with Delay JavaScript settings.
    • Improved: Grok web and image searches start faster and search for the key words of the question.
    • Fixed: Replies that reach the length limit end with a short note instead of stopping mid-sentence, also with AI Tools, and Claude can write longer replies.
    • Fixed: Image generation in the chat works again with Gemini and uses GPT Image 2.5 Flare with OpenAI, ahead of GPT Image 1 retiring on October 23, 2026. The Content screen uses the image model you choose.
    • Fixed: With delayed or on-interaction script loading, the chat waits for Complianz consent as it does on normal pages.
    • Fixed: The chatbot keeps working when the Moderation add-on is active and the Pro licence is not, and Moderation bans match visitors behind Cloudflare and guests who gave a banned email.
    • Fixed: Ratings and feedback are deleted and exported with their conversation, search replies are saved to the transcript once, and guest IP addresses show in transcript details.
    • Fixed: Code blocks keep all their code and Copy copies exactly what is shown, the Leads list is readable on laptop screens, code block labels are easier to read on light themes, and the auto-retry switch stays off when turned off.

  4. MxChat 3.2.25

    • New: License holders can turn MxChat Pro on or off for each of their sites from the Licenses page of their mxchat.ai account. The site checks once a day, warns its admin and pauses Pro after three days unless it is turned back on.
    • New: A switch on the Page Cache Compatibility card turns off the automatic page-cache clear that runs when MxChat settings are saved, for large sites that manage their own cache.
    • New: Re-import affected entries can also refresh entries whose page was edited after it was imported, for sites that keep auto-sync off.
    • Fixed: Gemini chatbots no longer stop with “The response format was unexpected” when Google’s safety filter flags an ordinary answer. The filter now blocks only high-risk content.
    • Fixed: A visitor message with a web link and a word like “read” or “documentation” is no longer treated as a PDF and answered with a PDF error.
    • Fixed: MxChat’s chat styles no longer change the rest of your site, such as code blocks on your own pages turning black.
    • Fixed: On a page with an embedded chatbot, the floating bubble of the same bot is no longer added. It opened but could not send messages.
    • Fixed: Links in the chat widget are correct for a www address at the end of a paragraph and for a web address written in bold.
    • Fixed: WordPress Content import and auto-sync now include text from synced patterns and WoodMart HTML Blocks. Save or import a page again to update its entry.
    • Fixed: Saving a product no longer removes its Custom Product Tabs text from the chatbot’s knowledge.
    • Fixed: With a Gemini 3 model, web search and image search use the key words of the visitor’s request instead of the whole sentence.
    • Fixed: The Knowledge list’s PDF filter shows PDF pages again, and uploaded files appear by name instead of a View Source link that led nowhere.
    • Fixed: Opening the Leads tab no longer writes a notice to the debug log when WordPress debugging is on.
    • Improved: Gemini 2.5 Pro, Flash and Flash-Lite are marked as no longer offered to new Google API users, and the admin is warned by name if Google refuses the selected model.
    • Improved: A chatbot still set to a retired AI model now uses that model’s current replacement on every request.
    • Improved: The Script Loading card explains that a cache plugin’s Delay JavaScript setting must leave MxChat’s chat script out, or the chatbot stays hidden until the visitor interacts.

  5. MxChat 3.2.24

    • Fixed: A live-agent message Slack refuses (deleted or archived channel, revoked token) now tells the visitor instead of being reported as sent, ends the handoff so the chatbot takes over again, and shows the error on the Slack tab. Slack messages also no longer begin with stray characters.
    • Fixed: The WordPress personal data export now includes visitors who filled in the pre-chat form without chatting (email, name and consent record).
    • Fixed: A media file whose knowledge entry was deleted is indexed again the next time it is saved, instead of being skipped as unchanged.
    • Fixed: View indexed content now works for entries whose address carries encoded characters (non-Latin pages, addresses with spaces) and for manually added Pinecone entries, and shows a long entry in full.
    • Fixed: Links in chatbot replies now open in Safari on iPhone, iPad and Mac when Open Links in New Tab is on. Safari’s pop-up blocker was silently blocking them.
    • Fixed: The video card for a matching YouTube entry now appears on streamed answers from GPT-5 and GPT-6 models. Before, it needed AI Tools on or streaming off.
    • Fixed: A chat message containing a document link no longer fails on chatbots embedded with MxChat Anywhere.
    • Fixed: Answers name an uploaded document or PDF in plain text instead of showing its file name as a link, including files uploaded before this update.
    • Fixed: A PDF added through PDF Upload keeps its file name. Its pages are listed under the file, answers name the file and page, and uploading the same PDF again replaces its pages instead of adding a second copy. PDFs uploaded earlier are left as they are: upload once more for a clean set, then del…
    • Fixed: On a Pinecone document index, part numbers and codes that start with digits are found again, such as an order number like 5872EF or 4471AB12 for a part printed 4471-AB-12. Entries written earlier keep their old codes: run Migrate again or re-save the entry to refresh them.
    • Fixed: Sites embedding through a Custom Provider can add knowledge through the REST API without an OpenAI key, and a Pinecone document index is sized from the custom model’s real dimension. The Knowledge listing, search, action phrases and import notice follow the custom model too.
    • New: Document Upload and PDF Upload each have an optional Source URL field. Give the page the file lives on and answers drawn from it link to that page.
    • New: WooCommerce product tags are included in each product’s knowledge-base text, beside its categories. Re-save or re-import a product to pick them up.
    • New: Re-import affected entries, on the Knowledge screen and as the wp mxchat kb-reimport command. It finds entries that earlier versions stored with currency signs or accents missing, HTML character codes left in, or right-to-left PDF text saved backwards. It shows the list first and changes nothi…
    • Improved: Delete old index now refuses while Multi-Bot bots still use that index, and names them.
    • Improved: With Multi-Bot on a site that keeps knowledge in the WordPress database, the Knowledge screen now says the content is shared with all bots. Per-bot separation needs Pinecone or an OpenAI Vector Store.
    • Improved: Admin polish. The warning status pill is styled, the importer’s selection counter has its spacing, the Pinecone card’s status follows the host you typed, and namespace lists keep a stable order.

  6. MxChat 3.2.23

    • Fixed: Pressing Enter with an empty message no longer locks the chat input. The box stays usable, and only a real message disables it while the reply is generated. A message made only of spaces or blank lines now counts as empty too, so nothing is sent to the model for it.
    • Improved: Pages imported from a web address or a sitemap now remember their own page title (from the page’s title tag or its social-sharing title, with the site name trimmed off). Add-ons that name a source, such as MxChat AI Search’s source cards, can show the real title instead of a guess made fr…
    • Improved: On a Pinecone document index, part numbers typed without their spaces, hyphens or dots now find the right record. R165321320 reaches the records that describe R1653 213 20 or the R1653 family, in any letter case, and an exact match always ranks above a family match. Records copied or adde…
    • Improved: Migrate now copies your vector index into the document index while the chatbot keeps answering from the vector index. Nothing switches until you save the Pinecone settings, and saving is held back while a copy into that index is still running unless you tick Switch anyway. An interrupted…
    • Fixed: Create index for me no longer fails with an HTTP 409 when the Index Name field still holds your vector index’s name. Switching to the document index type suggests a free name ending in -docs, an existing name is refused in plain words with a suggestion filled in, and creating or checking an…
    • New: The GPT-6 models are in the model picker: GPT-6 Astra (recommended), GPT-6 Sol and GPT-6 Luna. They work for chat, the content generator and the editor assistant, and the request settings each one requires are applied automatically.
    • Improved: When AI Tools are on, the model is now told how many tool calls it may make for a message, is told how many remain after each round, and is asked for a single call when only one is left. Before, a model could request several lookups at once when only one was allowed, have the rest refused…
    • Fixed: Delete old index now checks every namespace of the old vector index, not just the one copied last. It refuses, and names the namespace, while any namespace with records has no finished copy whose count matches. Before, a site with several namespaces (one per bot, or a namespace from an earli…
    • New: Migrate can copy every namespace of the vector index in one run. When the source index holds records in other namespaces, the Pinecone card offers a Copy every namespace switch beside Migrate. Each namespace is copied into the same-named namespace of the document index, so every bot keeps read…
    • Older releases are archived in the plugin’s full version history on WordPress.org.

  7. MxChat 3.2.22

  8. MxChat 3.2.21

  9. MxChat 3.2.20

  10. MxChat 3.2.19

  11. MxChat 3.2.18

  12. MxChat 3.2.17

  13. MxChat 3.2.16

  14. MxChat 3.2.15

  15. MxChat 3.2.14

  16. MxChat 3.2.13

  17. MxChat 3.2.12

  18. MxChat 3.2.11

  19. MxChat 3.2.10

  20. MxChat 3.2.9

  21. MxChat 3.2.8

  22. MxChat 3.2.7

  23. MxChat 3.2.6

  24. MxChat 3.2.5

  25. MxChat 3.2.4

  26. MxChat 3.2.3

  27. MxChat 3.2.2

  28. MxChat 3.2.1

  29. MxChat 3.2.0

  30. MxChat 3.1.9

  31. MxChat 3.1.8

  32. MxChat 3.1.7

  33. MxChat 3.1.6

  34. MxChat 3.1.5

  35. MxChat 3.1.4

  36. MxChat 3.1.3

  37. MxChat 3.1.2

  38. MxChat 3.1.1

  39. MxChat 3.1.0

  40. MxChat 3.0.9

  41. MxChat 3.0.8

  42. MxChat 3.0.7

  43. MxChat 3.0.6

  44. MxChat 3.0.5

  45. MxChat 3.0.4

  46. MxChat 3.0.3

  47. MxChat 3.0.2

  48. MxChat 3.0.1

  49. MxChat 3.0.0

  50. MxChat 2.6.4

202572 releases

  1. MxChat 2.6.3

  2. MxChat 2.6.2

  3. MxChat 2.6.1

  4. MxChat 2.6.0

  5. MxChat 2.5.9

  6. MxChat 2.5.8

  7. MxChat 2.5.7

  8. MxChat 2.5.6

  9. MxChat 2.5.5

  10. MxChat 2.5.4

  11. MxChat 2.5.3

  12. MxChat 2.5.2

  13. MxChat 2.5.1

  14. MxChat 2.5.0

  15. MxChat 2.4.9

  16. MxChat 2.4.8

  17. MxChat 2.4.7

  18. MxChat 2.4.6

  19. MxChat 2.4.5

  20. MxChat 2.4.4

  21. MxChat 2.4.3

  22. MxChat 2.4.2

  23. MxChat 2.4.1

  24. MxChat 2.4.0

  25. MxChat 2.3.9

  26. MxChat 2.3.8

  27. MxChat 2.3.7

  28. MxChat 2.3.6

  29. MxChat 2.3.5

  30. MxChat 2.3.4

  31. MxChat 2.3.3

  32. MxChat 2.3.2

  33. MxChat 2.3.1

  34. MxChat 2.3.0

  35. MxChat 2.2.9

  36. MxChat 2.2.8

  37. MxChat 2.2.7

  38. MxChat 2.2.6

  39. MxChat 2.2.5

  40. MxChat 2.2.4

  41. MxChat 2.2.3

  42. MxChat 2.2.2

  43. MxChat 2.2.1

  44. MxChat 2.2.0

  45. MxChat 2.1.9

  46. MxChat 2.1.8

  47. MxChat 2.1.7

  48. MxChat 2.1.6

  49. MxChat 2.1.5

  50. MxChat 2.1.4

  51. MxChat 2.1.3

  52. MxChat 2.1.2

  53. MxChat 2.1.1

  54. MxChat 2.1.0

  55. MxChat 2.0.9

  56. MxChat 2.0.8

  57. MxChat 2.0.7

  58. MxChat 2.0.6

  59. MxChat 2.0.5

  60. MxChat 2.0.4

  61. MxChat 2.0.3

  62. MxChat 2.0.2

  63. MxChat 2.0.1

  64. MxChat 2.0.0

  65. MxChat 1.6.5

  66. MxChat 1.6.4

  67. MxChat 1.6.3

  68. MxChat 1.6.2

  69. MxChat 1.6.1

  70. MxChat 1.6.0

  71. MxChat 1.5.9

  72. MxChat 1.5.8

202436 releases

  1. MxChat 1.5.7

  2. MxChat 1.5.6

  3. MxChat 1.5.5

  4. MxChat 1.5.4

  5. MxChat 1.5.3

  6. MxChat 1.5.2

  7. MxChat 1.5.1

  8. MxChat 1.5

  9. MxChat 1.4.6

  10. MxChat 1.4.5

  11. MxChat 1.4.4

  12. MxChat 1.4.3

  13. MxChat 1.4.2

  14. MxChat 1.4.1

  15. MxChat 1.4

  16. MxChat 1.3

  17. MxChat 1.2.1

  18. MxChat 1.2

  19. MxChat 1.1.91

  20. MxChat 1.1.9

  21. MxChat 1.1.8

  22. MxChat 1.1.7

  23. MxChat 1.1.6

  24. MxChat 1.1.5

  25. MxChat 1.1.4

  26. MxChat 1.1.3

  27. MxChat 1.0.5

  28. MxChat 1.1.2

  29. MxChat 1.1.1

  30. MxChat 1.1.0

  31. MxChat 1.0.11

  32. MxChat 1.0.10

  33. MxChat 1.0.9

  34. MxChat 1.0.8

  35. MxChat 1.0.7

  36. MxChat 1.0.6

Questions about MxChat’s releases

What is the latest version of MxChat?

3.2.28, released on Nov 6, 2026 (2 days ago). It’s tested up to WordPress 7.1.4 and needs PHP 7.4 or later.

How often is MxChat updated?

61 releases in the last year, about one every 6 days. Its first release on WordPress.org dates from September 2024.

Know when your rivals ship

Track your plugin and its rivals: TrackWP sees each new release in the nightly snapshot and alerts you, with what it did to their rank.

Start free

Release dates come from the day each version’s folder was made in MxChat’s WordPress.org SVN repository; the notes from its readme’s changelog, as its authors wrote it.