Skip to content

Releases: MindWorkAI/AI-Studio

Release v0.9.18

10 Nov 00:14
6928869
Compare
Choose a tag to compare
Release v0.9.18 Pre-release
Pre-release

v0.9.18, build 193 (2024-11-09 21:10 UTC)

  • Added new Anthropic model claude-3-5-heiku-20241022 as well as the alias claude-3-5-heiku-latest.
  • Added Groq as a provider option.
  • Added Google Gemini as a provider option.
  • Improved model display while configuring a provider when the provider supports display names.
  • Improved the bias catalog by cleaning up some artifacts.
  • Refactored the runtime code as preparation for the integration of RAG (integrating your data and files in AI Studio).

Virus scans

Release v0.9.17

03 Nov 14:36
b9416e6
Compare
Choose a tag to compare

Release v0.9.16

03 Nov 01:04
629613f
Compare
Choose a tag to compare

v0.9.16, build 191 (2024-11-02 22:04 UTC)

  • Added our second financial contributor: Thanks peerschuett for supporting AI Studio financially.
  • Added multiple workspace display options: toggle the workspace overlay (this is the previous behavior), toggle the workspace sidebar, or permanently display the workspace sidebar.
  • Improved the layout of the app window, for example, in full-screen mode on large or high-resolution monitors.
  • Fixed the "send to" issue, where you could select the bias-of-the-day assistant as destination.
  • Fixed the appearance of the "no diff" message for text diffs when the dark mode is enabled.

Virus scans

Release v0.9.15

28 Oct 18:14
c1d5be6
Compare
Choose a tag to compare

v0.9.15, build 190 (2024-10-28 15:04 UTC)

  • Added bias-of-the-day assistant: the assistant will provide a cognitive bias every day. The bias is explained in detail, and the user can ask questions about it.
  • Upgraded MudBlazor to v7.14.0.
  • Upgraded MudBlazor.Markdown to v7.14.0.
  • Upgraded tokio to v1.41.
  • Upgraded keyring to v3.6.
  • Upgraded arboard to v3.4.1.
  • Upgraded HtmlAgilityPack to v1.11.70.

Virus scans

Release v0.9.14

18 Oct 12:17
0d6ccbd
Compare
Choose a tag to compare

Release v0.9.13

07 Oct 15:41
a19a6b2
Compare
Choose a tag to compare

Release v0.9.12

15 Sep 23:47
776fa8a
Compare
Choose a tag to compare

v0.9.12, build 187 (2024-09-15 20:49 UTC)

  • Added a job posting assistant to the business category.
  • Improved text diff viewer by line-through deleted text.
  • Improved grammar & spelling check and rewrite assistants by showing the final text as well.
  • Fixed the dark theme for the text diff viewer, e.g., in the grammar check assistant.
  • Fixed margin-related issue in the ConfigurationText component.
  • Fixed chat conversion of job posting, legal check, and my task assistants.
  • Refactored the ConfigurationText component to debounce the input field to prevent unnecessary configuration updates. The component now also supports multiline text.

Virus scans

Release v0.9.11

15 Sep 13:32
b99866b
Compare
Choose a tag to compare

v0.9.11, build 186 (2024-09-15 10:33 UTC)

  • Added an option to enforce a minimum confidence level throughout the entire app.
  • Added options to enforce minimum confidence levels for each assistant individually.
  • Added color themes (dark & light) to the app settings. The theme can be synchronized with the system theme.
  • Added a tooltip to the confidence card button.
  • Renamed the Providers enum to LLMProviders for better clarity. Renamed also all dependent variables and methods.

Virus scans

Release v0.9.10

12 Sep 23:59
5608b48
Compare
Choose a tag to compare

v0.9.10, build 185 (2024-09-12 20:52 UTC)

  • Added support for OpenAI's o1 models. The 01 models presented today are initially limited to tier 5 accounts. This means that only individuals who have invested at least $1,000 in the OpenAI API and whose account is at least 30 days old can see and use 01 models. Once your OpenAI account has been activated for the new o1 models, you can select these models from the model list in the provider settings.
  • Added another source link for the OpenAI's confidence card.

Virus scans

Release v0.9.9

12 Sep 00:14
bf01fbd
Compare
Choose a tag to compare

v0.9.9, build 184 (2024-09-11 21:10 UTC)

  • Added confidence information to LLM providers.
  • Added the possibility to choose from predefined confidence schemes.
  • Added the possibility to define your own confidence scheme.
  • Added an option to show/hide confidence information in the app.
  • Added a visual indicator for confidence in the chat and all assistants.
  • Added a confidence card to the chat and all assistants.

Virus scans