Introducing the new Demo Playground

#57
by hdrjt - opened
UltimaX Intelligence org

Today, I am releasing the Demo Playground with a refreshed design and added support for multimodal capabilities.

It continues to run on the GPT-5 model.

https://huggingface.co/spaces/umint/ai
https://umint-ai.hf.space

UltimaX Intelligence org
edited 29 days ago

Update October 15, 2025

  • Switched to a limited model for long context memory.

  • Fixed UI lag issues when rendering various types of LLM responses.

UltimaX Intelligence org
edited 29 days ago

Switch back to regular model

UltimaX Intelligence org

Update October 16, 2025

  • The maximum input limit for demo use is now set to 150 characters.

  • Added swipe-to-close support for the sidebar, specifically for mobile and smartphone modes.

where do you get the ai api's? pollinations doesn't have gpt-4.1...

UltimaX Intelligence org

Update October 19, 2025

  • Completely switched to a limited model.

  • Refactor backend logic.

UltimaX Intelligence org

Update October 19, 2025

  • Added support for multiple providers.

  • Added support for multiple models.

  • Introduced a new model for the Demo Playground.

    • Normal models:

      1. openai/gpt-5-chat-latest
      2. openai/gpt-4.1-nano-2025-04-14
      3. openai/o4-mini-2025-04-16
    • Limited models:

      1. limited/gpt-5-chat-latest
      2. limited/gpt-4.1-2025-04-14
      3. limited/gpt-4o-2024-11-20

    Click here to view full details about the limited models.

  • Demo Playground updated to 1.0.4 version.

UltimaX Intelligence org

Update October 22, 2025

  • Optimized React components
  • Optimized Next.js components
  • Set maximum upload size limit for images
  • Set limit for the number of images per upload
  • Optimized layout for mobile and smartphone use
  • Enhanced the security of the Demo Playground ecosystem
  • Integrated Cloudflare Turnstile for human verification
  • Added informational pop-ups for specific scenarios
  • Updated Next.js to 16.0.0 version (latest)
  • Updated Demo Playground to 1.0.5 version

Enjoying this project and finding it helpful? Consider making a donation.

Your contribution will help make the ecosystem even more powerful in the future.

UltimaX Intelligence org

Update October 23, 2025

  • Optimized Turbopack configuration to align with updates in Next.js 16.0.0 version.

  • Fixed an issue where model switching occasionally remained on the default model when triggering example prompts.

  • Minor improvements to the verification layout.

  • Added DeepSeek R1 (0528) model.

  • Updated Demo Playground to 1.0.6 version.

Enjoying this project and finding it helpful? Consider making a donation.

Your contribution will help make the ecosystem even more powerful in the future.

UltimaX Intelligence org

Update October 25, 2025

  • Frontend:

    • Redesigned the model selector interface
    • Made the chat area fully responsive to full screen
    • Added expandable and collapsible header functionality
    • Fixed minor bugs in the input form
  • Backend:

    • Updated Next.js compatibility for Hugging Face Spaces
    • Optimized initial web rendering for faster load times
  • Added openai/gpt-oss-20b model

Enjoying this project and finding it helpful? Consider making a donation.

Your contribution will help make the ecosystem even more powerful in the future.

UltimaX Intelligence org

Update November 4, 2025

  • Updated Next.js to 16.0.1 version (latest)
  • Bumped Demo Playground to 1.0.7 version
  • Fixed UI lag and performance delays
  • Fixed minor issues in the multimodal input form
  • Improved smooth streaming response
  • Updated model labels for clarity
  • Show all error logs to the frontend chat interface
  • Added MiniMaxAI/MiniMax-M2 model

Enjoying this project and finding it helpful? Consider making a donation.

Your contribution will help make the ecosystem even more powerful in the future.

hdrjt pinned discussion

Okay, great, so why was main Ultima Intelligence closed down?

UltimaX Intelligence org

Okay, great, so why was main Ultima Intelligence closed down?

https://huggingface.co/spaces/umint/ai/discussions/63#68fa10c7718bf18d0528b1e3

UltimaX Intelligence org

Update November 9, 2025

  • Accelerated the streaming process while ensuring smooth performance and minimal waiting time.

    --> Dependent on the server provider’s speed.

  • Implemented adaptive chunking for streaming responses.

  • Minor bug fixes for Stop/Cancel/AbortController functionality.

  • Added target="_blank" to markdown and HTML links to prevent inline web rendering when links are clicked.

  • Implemented lazy loading for the syntax highlighter.

  • Added several new models, click here for details.

  • Bumped Demo Playground to 1.0.8 version.

Enjoying this project and finding it helpful? Consider making a donation.

Your contribution will help make the ecosystem even more powerful in the future.

Sign up or log in to comment