Support Forum

Hi, thank you for the screenshot and the update about desktop. We're sorry for the display problem.

We found a regression in today's update that stopped chat text from wrapping, causing conversations to extend past the right edge on both mobile and desktop. The fix is now live, and we've verified text wrapping and lay

Update: on PC version, scale's wrong, too ('cut' on the right side) even without F12 enabled (sorry, can't add a screen)

Its back to normal, thanks

Hi Alnor,

We've fixed today's issue that prevented previous chats from appearing. We understand how worrying it is when the conversations containing your prompts and image/video results seem to disappear, and we're sorry for causing that concern.

The problem was with loading chat history after our update. We've verif

Much obliged.

Hi Jei Are,

We've fixed the issue that caused “No previous chats,” “You have no update permissions,” and “Duplicate Data Entered.” Today's update introduced a problem on our side that prevented chat history from loading and messages from saving correctly. We're sorry for the disruption and the worry it caused.

We've

Hi,

Thanks for reaching out, and sorry for the inconvenience.

1. Image Display Issue:
   The site was not down for maintenance. A recent update introduced a bug in the public link resolution path, which caused existing generated images to temporarily return 404 errors. All image files remained completely safe on storage

I shouldn't say 'gone' I should have said 'refuse to load'. There's a broken image icon where the .jpgs should be.

Hi! Thanks for reaching out and sharing this feedback.

Currently, GizAI doesn't support adding custom labels or tags directly to individual image cards inside a chat conversation stream. When creating many iterations in a single session, we understand how keeping track of which version is which can become difficult.

Ok . ty. i will try this one

Hi Alnor,

Thank you for reaching out! When a video generator doesn't follow a prompt despite rewriting it through GPT, Claude, or Qwen, the issue is almost always architectural rather than prompt wording.

Here is why this happens and the professional workflow to achieve your exact vision:

1. Why DiT Models Strug

Hi! Sorry for the disruption, and thank you for explaining it so clearly. We've fixed the text jumping on hover and restored the Refresh button under Billing → Rate limits. These changes are now live.

We also checked that opening Edit keeps the chat in place without reloading. Please refresh GizAI once and try again.

Thanks for clarifying! I refreshed the page, switched to GPT6 Astra, and everything is working as expected ,its awesome! thanks

Thanks for reaching out, and sorry for the confusion. GPT-6 Astra has replaced GPT-5.6 Sol as an Included model, with the same usage allowance for your plan.

Please refresh the page and select GPT-6 Astra under Included models to use that allowance. Sol is still available separately as a usage-billed model, which is w

Sorry you've run into another issue with GPT Image 2, and thank you for reporting it. We've fixed the "File id is required" error, and the update is now live. Please refresh the page and try again. Let us know if it still occurs.

hello Gizai team, thanks for the tip, i will try all of this tips and will be back here for update
,

Hi Alnor!

Thank you for the detailed report! We investigated your exact 15-second generation runs with LTX-2.5 to see why the output shifts at the 7–8 second mark.

Here is what is happening under the hood and how to achieve seamless 15-second results:

Why the Drift Happens Around Second 7–8 in LTX-2.5:

1. **Init