Skip to content

Releases: Krzyzyk33/CMDAI

v1.1.2beta

18 Mar 17:17
15ff4e9

Choose a tag to compare

v1.1.2beta

New mode

  1. Debug mode
  2. Analityst mode

New UI part

  1. ◆ CHAT
  2. ◆ PLAN
  3. ◆ CODE
  4. ◆ DEBUG
  5. ◆ ANALYST

Terminal command run

  1. AI can run terminal comand

New /ide integration

  1. Cursor
  2. Intelij idea

Fix bugs

  1. Ide integration bugs
  2. UI bugs

AND MUCH MUCH MORE

Please report bugs

v1.1.1beta (CODE/PLAN mode)

02 Mar 19:49
2cc71d9

Choose a tag to compare

This is the largest update to CMDAI so far, introducing new modes! We've focused on enhancing usability and adding powerful tools for AI interaction. Please test thoroughly and report any bugs in the Issues section – your feedback is crucial!


🔄 New Modes

  1. Code Mode: Uses the file generated by Plan Mode to create the app. This allows seamless code execution based on planned logic.
  2. Plan Mode: Generates a detailed plan for Code Mode, helping structure complex tasks before implementation.

✨ New Functions

  1. Real-Time Model Activity Visibility: Now you can see what the model is doing in real-time (e.g., thinking, analyzing, etc.). This provides better transparency during operations.
  2. Writing Area: Added a dedicated space for writing with the model.
  3. Once you have installed it you can open the application from the terminal just write.
CMDAI launch 

⌨️ Commands

  1. Slash Prefix Requirement: From now on, commands only work when prefixed with /. We're still adding more commands in upcoming updates, as not all are fully implemented yet. Sorry for the inconvenience!

📦 Installation, Model Loading, and Code Execution

  1. Install CMDAI easily and load your GGUF models with simple terminal commands.
  2. Enhanced code execution support for smoother integration with your workflows.

🐞 Bug Reporting

  1. This major update may have some rough edges – please report any bugs or issues in the [GitHub Issues] (https://github.com/Krzyzyk33/CMDAI/issues) section. Your reports help us improve!
  2. Thank you for using CMDAI! Star the repo if you like it, and stay tuned for more updates. 🌟dowolad app in my GitHub repository (https://github.com/Krzyzyk33/CMDAI/releases/tag/v1.3.0)
  3. This is the largest update to CMDAI so far, introducing
    new modes, features, and commands! We've focused on enhancing usability and adding powerful tools for AI interaction. Please test thoroughly and report any bugs in the Issues section – your feedback is crucial!

CMDAI x VideoHub

27 Feb 16:04
3550e71

Choose a tag to compare

🚀 CMDAI x Video Hub Integration Update (v1.3 Beta)

  1. Video Now on Video Hub!
    All CMDAI demos, tutorials, and videos will be published on Video Hub – a new video management app. Easy browsing, tagging, and playback for local AI videos.

  2. New Video Comparison
    Check out the latest demo: LM Studio vs. Ollama vs. CMDAI (RAM usage, features, etc.).

VideoHub

  1. Collaboration: CMDAI x Video Hub
    All videos on VideoHub starting today!!!

Go to My Repositories

  1. CMDAI
  2. VideoHub

Feedback? Give me an issue or a star! 🌟

v1.1.0

23 Feb 18:11
1163172

Choose a tag to compare

Update 1.1.0

Released: 2026-02-23

Fix Bugs

  1. Fix load model (now worker progres).
  2. Fix UI.
  3. Added fallback for ./models permissions issue.
  4. Safer charging parameters (including fallback without mmap) to reduce the risk of hang on some models.
  5. Aliasy modeli są teraz ładowane dynamicznie z models/

Showing chat history to model (BETA)

  1. Now the model can read the histories of a given session without creating additional files.

New model generating UI

  1. Now model when generating answer there is a new look.

Model change during conversation

  1. You can now change your model while chatting with another

Detecking python if availabe

  1. There is also an auto-attempt to run the app on Python 3.11, if available.

Removing the token limit

  1. There is no limit now!!!

v1.1.0beta

20 Feb 15:11
5a322cf

Choose a tag to compare

Update 1.1.0beta

Released: 2026-02-20

Model Loading Improvements

  1. Added loading spinner for better visual feedback.
  2. Reduced console noise during model initialization.
  3. Fixed several loading‑related bugs.

Help UI Improvements

  1. Improved layout and readability of the help screen.
  2. Cleaned up formatting and spacing for better clarity.

Command Changes

  1. Removed quick commands to simplify the command structure.

Repository change

  1. Readme update
  2. New files ROADMAP.md and FAQ.md
  3. New file assets
  4. Demo coming soon

CMD LOCAL AI v1.0.0 – First Public Release

19 Feb 16:37
3b9d859

Choose a tag to compare

CMD LOCAL AI v1.1.0

What’s new

  • Added improved terminal chat flow with AI> prompt.
  • Added spinner during response generation.
  • Added response stats line with token count and generation time.
  • Improved output filtering to show final answer only.
  • Added safer handling for model loading retries and timeouts.
  • Updated docs (README.md, index.html) with Polish and English sections.

Highlights

  • Local-first GGUF model runner (llama-cpp-python)
  • CLI + Ollama-compatible API (/tags, /generate, /chat, /pull, /version)
  • Cleaner chat UX and more predictable answers

More updates coming soon.