Skip to content

Releases: MindWorkAI/AI-Studio

Release v26.2.1

01 Feb 19:40
@github-actions github-actions
v26.2.1
e65874d
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v26.2.1 Latest
Latest

v26.2.1, build 233 (2026-02-01 19:16 UTC)

  • Added the ability to individually configure the minimum confidence level, standard profile, and default provider for each policy in the Document Analysis Assistant (in preview).
  • Added support for defining document analysis policies (in preview) by configuration plugins, enabling centralized management of policies across entire departments or organizations.
  • Added an option to hide the policy definition in the Document Analysis Assistant (in preview) when exporting and distributing that policy by a configuration plugin in organizations, making it easier for users to use.
  • Added the policy export functionality to the Document Analysis Assistant (in preview). You can now export policies as Lua code for a configuration plugin to distribute the policy across your organization.
  • Improved the error checking & logging behavior when the installed PDFium version did not meet the minimum required version.
  • Improved the logging behavior when parsing profile tables from configuration plugins.
  • Fixed a bug where the global minimum confidence level was not being applied to the assistants.
  • Upgraded the Document Analysis Assistant and the voice recording with transcription feature (both in preview) from the prototype to the beta state. Both features are now completely implemented and are undergoing a deeper testing phase in preparation for release.

Virus scans

Assets 23

Release v26.1.2

25 Jan 14:29
@github-actions github-actions
v26.1.2
09f93c6
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v26.1.2

v26.1.2, build 232 (2026-01-25 14:05 UTC)

  • Added the option to hide specific assistants by configuration plugins. This is useful for enterprise environments in organizations.
  • Added the current date and time to the system prompt for better context in conversations. Thanks Peer peerschuett for the contribution.
  • Added the ability to control the voice recording with transcription (in preview) by using a system-wide shortcut. The shortcut can be configured in the application settings or by using a configuration plugin. Thus, a uniform shortcut can be defined for an entire organization.
  • Added error handling for the context window overflow, which can occur with huge file attachments in chats or the document analysis assistant.
  • Added Rust failure detection to the .NET server. This is helpful in order to handle critical failures and shut down the application gracefully.
  • Improved the error handling for model loading in provider dialogs (LLMs, embeddings, transcriptions).
  • Improved the microphone handling (voice recording & transcription preview) so that all sound effects and the voice recording are processed without interruption.
  • Improved the handling of self-hosted providers in the configuration dialogs (LLMs, embeddings, and transcriptions) when the host cannot provide a list of models.
  • Improved the document analysis assistant (in preview) by allowing users to send results to a new chat to ask follow-up questions. Thanks to Sabrina Sabrina-devops for this contribution.
  • Improved the developer experience by detecting incorrect CPU architecture metadata when checking and installing the Pandoc dependency.
  • Improved the error messages for failed communication with AI servers.
  • Improved the error handling for the enterprise environment service regarding the communication with our Rust layer.
  • Fixed a bug in the document analysis assistant (in preview) where the first section containing the policy definition did not render correctly in certain cases.
  • Fixed a logging bug that prevented log events from being recorded in some cases.
  • Fixed a bug that allowed adding a provider (LLM, embedding, or transcription) without selecting a model.
  • Fixed a bug with local transcription providers (voice recording & transcription preview) by handling errors correctly when the local provider is unavailable.
  • Fixed a bug with local transcription providers (voice recording & transcription preview) by correctly handling empty model IDs.
  • Fixed a bug affecting the transcription preview: previously, when you stopped music or other media, recorded or dictated text, and then tried to resume playback, the media wouldn’t resume as expected. This behavior is now fixed.
  • Fixed a rare bug that occurred when multiple threads tried to manage the same chat thread. Thanks Sabrina Sabrina-devops for identifying and reporting this issue.
  • Fixed a bug that prevented text localization from certain source code files under specific conditions.
  • Upgraded to Rust 1.93.0
  • Upgraded to .NET 9.0.12
  • Upgraded dependencies

Virus scans

Loading

Release v26.1.1

11 Jan 16:21
@github-actions github-actions
v26.1.1
ab76180
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v26.1.1

v26.1.1, build 231 (2026-01-11 15:53 UTC)

  • Added the option to attach files, including images, to chat templates. You can also define templates with file attachments through a configuration plugin. These file attachments aren’t copied—they’re re-read every time. That means the AI will pick up any updates you make to those files.
  • Added the option to use source code files in chats and document analysis. This supports software development workflows.
  • Added a preview feature that lets you record your own voice for transcription. The feature remains in development and appears only when the preview feature is enabled.
  • Added the option to configure transcription providers (for example, using Whisper models). As usual, there can be local as well as cloud models configured. This option is part of the transcription preview and remains hidden until the preview is activated or the feature gets released. Transcription providers can be configured through configuration plugins as well.
  • Added an option to the app settings to select a provider for transcribing dictated text. If no provider is selected, dictation and text transcription are disabled.
  • Added the option to configure embedding providers through a config plugin and distribute them within an organization.
  • Improved the app versioning. Starting in 2026, each version number includes the year, followed by the month. The last digit shows the release number for that month. For example, version 26.1.1 is the first release in January 2026.
  • Fixed a bug in the profile selection where the "Use no profile" entry could not be localized, causing English text to appear in languages such as German. This behavior has now been fixed.
  • Fixed a bug in the provider dialogs (LLMs, embeddings, and transcriptions) when editing a provider. In cases where an error had to be displayed, a non-localized message in English was used.
  • Fixed a very rare bug in the provider dialogs (LLMs, embeddings, and transcriptions) where a validation error appeared if the API key could not be read from the operating system, but the error did not clear after the user changed the API key.
  • Upgraded dependencies.

Virus scans

Loading

Release v0.10.0

31 Dec 14:36
@github-actions github-actions
v0.10.0
3b76b06
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v0.10.0

v0.10.0, build 230 (2025-12-31 14:04 UTC)

  • Added support for newer Mistral models (Mistral 3, Voxtral, and Magistral).
  • Added support for the new OpenAI model GPT 5.2.
  • Added support for OpenRouter as LLM and embedding provider.
  • Added support for multimodal processing (documents and images, for now) when the selected LLM supports it.
  • Added a description field to local data sources (preview feature) so that the data selection agent has more information about which data each local source contains when selecting data sources.
  • Added the ability to use file attachments (including images) in chat. This is the initial implementation of this feature. We will continue to develop this feature and refine it further based on user feedback. Many thanks to Sabrina Sabrina-devops for this wonderful contribution.
  • Improved the document analysis assistant (in preview) by adding descriptions to the different sections.
  • Improved the document analysis assistant (in preview) by allowing users to use images as input files in addition to documents.
  • Improved the document preview dialog for the document analysis assistant (in preview), providing Markdown, image and plain text views for attached files.
  • Improved the Pandoc handling for the document analysis assistant (in preview) and file attachments in chat. When Pandoc is not installed and users attempt to attach files, users are now prompted to install Pandoc first.
  • Improved the ID handling for configuration plugins.
  • Improved error handling, logging, and code quality.
  • Improved error handling for the Microsoft Word export.
  • Improved the file reading, e.g. for the translation, summarization, and legal assistants, by performing the Pandoc validation in the first step. This prevents unnecessary selection of files that cannot be processed.
  • Improved the file selection for file attachments in chat and the assistant file loading by filtering out audio files. Audio attachments are not yet supported.
  • Improved the developer experience by automating localization updates in the filesystem for the selected language in the localization assistant.
  • Improved the file selection so that users can now select multiple files at the same time. This is useful, for example, for document analysis (in preview) or adding file attachments to the chat.
  • Fixed a bug in the local data sources info dialog (preview feature) for data directories that could cause the app to crash. The error was caused by a background thread producing data while the frontend attempted to display it.
  • Fixed a visual bug where a function's preview status was misaligned. You might have seen it in document analysis or the ERI server assistant.
  • Fixed a rare bug in the Microsoft Word export for huge documents.
  • Fixed a bug in the chat options that occurred when selecting default data sources. Under certain conditions, selecting data sources caused an error that required restarting the app. This preview-only feature (RAG preview) had not been released yet.
  • Fixed a bug in the chat template selection where the "No chat template" entry could not be localized, causing English text to appear in languages such as German. This behavior has now been fixed.
  • Fixed a workspace-related bug where disabling workspaces completely prevented access to the workspace settings. Workspace settings now appear next to chat settings in this case. Thanks to Paul PaulKoudelka for reporting.
  • Upgraded dependencies such as Rust, MudBlazor, and others.

Virus scans

Loading

Release v0.9.54

24 Nov 18:59
@github-actions github-actions
v0.9.54
311092e
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v0.9.54

v0.9.54, build 229 (2025-11-24 18:28 UTC)

  • Added a preview of the document analysis. Documents can now be uploaded and analyzed according to predefined rules. Structured output can also be configured using corresponding rules. This preview must be explicitly enabled in the app settings; otherwise, it remains hidden. Thanks, Peer (peerschuett), for your support.

Virus scans

Loading

Release v0.9.53

17 Nov 12:51
@github-actions github-actions
v0.9.53
8b0ee49
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v0.9.53

v0.9.53, build 228 (2025-11-14 13:14 UTC)

  • Added expert settings to the provider dialog to enable setting additional parameters. Also, additional parameters can be configured by configuration plugins for enterprise scenarios. Thanks to Peer (peerschuett) for this contribution.
  • Added the ability to export AI responses from the chat into Microsoft Word files. Thank you, Sabrina (Sabrina-devops), for your first contribution.
  • Added the ability to use documents as input for profile fields.
  • Added the ability to distribute profiles via configuration plugins in enterprise environments. Thanks, Paul (PaulKoudelka), for your first contribution.
  • Added the ability to preset an app-wide default profile using a configuration plugin in enterprise environments.
  • Added support for the new OpenAI GPT 5.1 models.
  • Improved profiles by removing their input limits.
  • Improved the file reading component to correctly verify the Pandoc installation and open the installation dialog when needed.
  • Upgraded dependencies.
  • Upgraded to Rust v1.91.1.
  • Upgraded to .NET v9.0.11.

Virus scans

Loading

Release v0.9.52

24 Oct 07:24
@github-actions github-actions
v0.9.52
37bd42e
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v0.9.52

v0.9.52, build 227 (2025-10-24 06:00 UTC)

  • Added a feature so that matching results from data sources (local data sources as well as external ones via the ERI interface) are now also displayed at the end of a chat. All sources that come directly from the AI (like web searches) appear first, followed by those that come from the data sources. This source display works regardless of whether the AI actually used these sources, so users always get all the relevant information.
  • Added the ability to manage the preview feature visibility and enabled preview features by using enterprise IT configurations.
  • Improved developer experience by detecting development environments and disabling update prompts in those environments.
  • Fixed an issue where external data sources using the ERI interface weren't using the correct source names during the augmentation phase.

Virus scans

Loading

Release v0.9.51

04 Sep 18:41
@github-actions github-actions
v0.9.51
f00722e
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v0.9.51

v0.9.51, build 226 (2025-09-04 18:02 UTC)

  • Added support for predefined chat templates in configuration plugins to help enterprises roll out consistent templates across the organization.
  • Added the ability to choose between automatic and manual update installation to the app settings (default is manual).
  • Added the ability to control the update installation behavior by configuration plugins.
  • Added the option for LLM providers to stream citations or sources.
  • Added support for citations to the chat interface. This feature is invisible unless an LLM model is streaming citations or sources.
  • Added the Responses API according to the OpenAI documentation. It is currently only used by OpenAI, but we could use the API for other providers as soon as someone offers it. This means that all text-based LLMs from OpenAI can now be used in MindWork AI Studio. For example, the Deep Research models for comprehensive research tasks.
  • Added support for web searches. Currently supported by some OpenAI models (e.g., GPT5, newer Omni models, Deep Research models) and Perplexity. Used sources are displayed visually in the chat interface.
  • Added support for the LLM provider Perplexity. It can be used for web searches, deep research, and general chat tasks. Thanks peerschuett for the contribution.
  • Improved memory usage in several areas of the app.
  • Improved plugin management for configuration plugins so that hot reload detects when a provider or chat template has been removed.
  • Improved the dialog for naming chats and workspaces to ensure valid inputs are entered.
  • Improved the dialog invocation by making parameter provision more robust.
  • Improved the text summarizer assistant by allowing users to specify important aspects & optimized the generated prompt.
  • Improved the OpenAI provider by supporting more models and capabilities.
  • Improved the German translation.
  • Improved the enterprise configuration of complex objects, such as chat templates: the corresponding Lua tables are now optional. This ensures existing enterprise configurations remain valid when new configuration options are added and don't require adjustments with every update.
  • Changed the configuration plugin setting name for how often to check for updates from UpdateBehavior to UpdateInterval.
  • Fixed a bug in various assistants where some text fields were not reset when resetting.
  • Fixed the input field header in the dialog for naming chats and workspaces.
  • Fixed a rare chat-related bug that could occur when a workspace was not created correctly. Thank you, Naomi, for reporting this issue.
  • Fixed a bug in the web content reader where a preconfigured LLM provider was not recognized, resulting in an error message.
  • Fixed another bug in the web content reader: the system incorrectly evaluated whether the content cleaning agent was preselected. Users had to first deactivate the agent and then reactivate it to make it work correctly.
  • Fixed a bug in the assistants for text summarization, translations, and legal check: the web content reader preferences were not being applied when resetting the assistant.
  • Fixed the bottom margin of code blocks in chat. The bottom margins were too small. This was particularly problematic when horizontal scrollbars were necessary.
  • Fixed a bug in assistants: the selected profile wasn't reset when resetting the assistant.
  • Upgraded dependencies.

Virus scans

Loading

Release v0.9.50

10 Aug 17:25
@github-actions github-actions
v0.9.50
28b470f
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v0.9.50

v0.9.50, build 225 (2025-08-10 16:40 UTC)

  • Added an option for chat templates to predefine a user input.
  • Added the ability to create chat templates from existing chats.
  • Added an enterprise IT configuration option to prevent manual addition of LLM providers in managed environments.
  • Added support for self-hosted LLMs using vLLM.
  • Improved the display of enterprise configurations on the about page; configuration details are only shown when needed.
  • Improved hot reloading on Unix-like systems when entire plugins were added or removed.
  • Improved the German translation.
  • Fixed a visual bug in all assistants that occurred when no profile was selected.
  • Upgraded .NET to v9.0.8
  • Upgraded MudBlazor to v8.11.0
  • Upgraded Rust to v1.89.0
  • Upgraded dependencies.

Virus scans

Loading

Release v0.9.49

02 Jul 12:35
@github-actions github-actions
v0.9.49
9c52812
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
Release v0.9.49

v0.9.49, build 224 (2025-07-02 12:12 UTC)

  • Added a library by Nils Kruthoff (nilskruthoff) that allows AI Studio to read PowerPoint files. Thanks, Nils, for that great contribution.
  • Added support for loading arbitrary document data into some assistants. This functionality replaces the previous PDF reading function and was contributed by Nils as well.
  • Improved the loading of some components that require data fetching, resulting in a more responsive UI.
  • Improved some awkward phrasings in English and German.
  • Improved the implementation of configuration plugins to enhance long-term maintainability.
  • Changed the timestamp display to use the local datetime format for the chats and assistants.
  • Fixed a bug when editing data sources that caused the selected retrieval process of an ERI data source to not load correctly.
  • Fixed a bug in the chat that caused the multi-line text input box to occupy too much space even after submitting a prompt.
  • Upgraded to Rust 1.88.0.
  • Upgraded MudBlazor to version 8.9.0.
  • Upgraded Pdfium to version 137.0.7215.0.
  • Upgraded other dependencies to the latest versions.

Virus scans

Loading