Home Creators Posts Import Register Favorites Logout

Downloads

Content

FramePack Studio, a popular fork of FramePack, installs in 1 click with pre-compiled libraries that support RTX 5000 series and older GPUs.

Automatic installer for the latest version (v0.4.0) on Windows, RunPod and Massed Compute, including RTX 5000 series support; running the Windows installer again updates it. You can reuse an existing FramePack models folder, and a big LoRA list for FramePack is linked.

1-Click to install popular FramePack Studio latest version into your Windows Computer with RTX 5000 and older GPUs support + RunPod + MC

Patreon exclusive posts index to find our scripts easily, Patreon scripts updates history to see which updates arrived to which scripts and amazing Patreon special generative scripts list that you can use in any of your task.

Join discord to get help, chat, discuss and also tell me your discord username to get your special rank : SECourses Discord

Please also Star, Watch and Fork our Stable Diffusion & Generative AI  GitHub repository and join our Reddit subreddit and follow me on LinkedIn (my real profile)

=======

Latest installer zip file : FramePack_Studio_v2.zip

  • FramePack Studio another popular fork of FramePack - not developed by me

  • Big list with LoRAs for Framepack https://fpsloralist.netlify.app/

  • Its repo : https://github.com/colinurbs/FramePack-Studio

    • If you want to use your existing previous models folder you can edit studio.py inside the FramePack-Studio folder after installation and change this to your previous folder where they are downloaded

    • e.g.:

      • os.environ['HF_HOME'] = r"Q:\FramePack_v1\FramePack\hf_download"

11 June 2025 v0.4.0

  • There have been a lot of updates to the app

  • Just run Windows_Install.bat and it will get updated to latest version

Windows Requirements

  • Python 3.10.11, FFmpeg, CUDA 12.8 or above, cuDNN 9.4 or above, C++ tools, MSVC and Git

  • If it doesn't work make sure to below tutorial and install everything exactly as shown in this below tutorial

  • https://youtu.be/DrhUHnYfwC0

Windows:

  • First watch above requirements tutorial and follow every step

    • Extract zip archive into a drive like c:/FramePack_v1

    • Longer directory path or space character or special character in folder paths causes issues

    • Double click and install with Windows_Install.bat

    • Then run with Windows_Start_App.bat

      • If you get model download error open Windows_Start_APP.bat with a notepad editor and make "HF_HUB_ENABLE_HF_TRANSFER=0"

Massed Compute (Recommend Cloud) :

RunPod (Cloud):

Related posts

All posts in this topic: AI Video Generation and Animation Apps · New to SECourses? Start Here: Best and Most Popular SECourses Posts

Files

Comments

DanO..

What's different? I see some stuff moved around and removed. FramePack F1 is a huge improvement for me. Thanks for getting that running.

SECourses

well this was requested i dont know what extra it has than ours :D

DanO..

Gotcha! It's missing LoRA strength for one thing. I'm sure those that wanted it are happy though.

koon grizzly

i edited the python file and must have closed without saving lol, so i got a full fresh install (again) i already installed the original framepack studio also, i like it because it has text to video and works pretty well, i tested a couple of prompts and much was followed But i haven't used it because it doesn't have a setting for FPS so it's way slower (for me with the setting at 4 seconds per video part) it also don't support save frames and other usefull options i got used to

koon grizzly

it has text to video and the queue page looks great with easy to find details of what was done. Other then that i don't know how to get 480p, it doesn't save frames, no intermediate video files, no fps,.... so i only used it for testing text to video

Art of 3D Rendering

thanks that is really cool can have both versions in one place!

Mordred Sillence

I had to do some serious editing to get it running on runpod. I hope you fix it.

koon grizzly

i meanwhile learned that it saves the parts in a temp folder and it has a feature i like : when i cancel the job it will give me whatever what was done already in a finished mp4 file. so let's say i cancel a video of 2 minutes when it has 30 seconds i will get a 30 second video file in output as 1 mp4. I have been testing the text to video feature more last night, made a couple of 2 minute videos, this is next level for me :-) it creates full 2 minute stories and switches to new scenes for new prompts without the need for any photo, only needs a good prompt.

Bertrand Samimi

Great job Why we cannot change framerate Now is all times 30 In Europa we use 25

Anthony

maybe you should be making your own autoinstallers then.

stphnvdb321

It's nice to use FramePack with text to video. I made a few tests and every video turned into a cartoonish style after some seconds. Is it just me? This didn't happen with the original FP.

SECourses

well i think text to video is just regular hunyuan not framepack

SECourses

well this app not developed by me :D you can open issue and request

SECourses

well this app not developed by me :D you can open issue and request

SECourses

sorry let me test and see if anything need fix

SECourses

thanks. i will add a new version too i saw today :D

SECourses

ye this is different not have everything we have :D

stphnvdb321

But you can use text to video with FramePack Studio up to 120s. In the original FP you had only image to video.

SECourses

hello just fixed with v2 there was an inaccurate torch version

Hipno

i removed my comment, i made a mistake when installing in massed compute. i was trying to run it from the thindrive instead of copying it over to the downloads folder in massed compute haha

SECourses

it generates a black image and uses it. adding to our app as well right now

SECourses

ah yes that is a common mistake :D

stphnvdb321

And Studio has also timed promts (like your app). But with every second of video, the quality gets worse. At least for me.

koon grizzly

when using animals i use this prompt : A realistic anthropomorphic (raccoon , lion,...) Gives real animals with human gestures every time If you want it to be the animal itself : a realistic anthropomorphic real life (raccoon , lion,...)

koon grizzly

+1 without change fps latent window size don't work to make videoparts longer so takes more time to create :(

SECourses

i also added to our app text prompt feature and custom resolution feature

Mordred Sillence

Thank you very much. I knew that you would find the problem.

koon grizzly

Anyone tested 2 minute video in framepack studio without timestamps ? (only one prompt ) Seems like it often restarts from starting photo. (F1 version) I put a couple in the queue before going to bed. (which got stuck by the time i woke up but it kept going in the background) Another thing i noticed is that you can see almost every second that it created a new part. (but maybe the reason is because i only used 20 steps to speed things up) Check the examples. https://youtu.be/tDDH-O2-6sw https://youtu.be/7NoB8l9dvTY

Goldo 27

they just had a new version out today with a lot of changes. Are you going to update it here too soon?

SECourses

my installer installs latest version always. run Windows_Install.bat and it will update to latest version . i just tried and it works. alone windows update bat fails since there is a newer requirement

puk

THX for the update. When I generate a video, no video appears in the preview window

SECourses

ye sometimes happens i dont know what exactly causes that

cool1

It seems good, and it has an option to remove the temporary files, though it doesn't have the "batch processing" option of the main Frame Pack AI, where you can get it to process a whole folder and ask it to create multiple videos for each.

cool1

Though it has trouble continuing overnight on my RTX3090 at 896x576. It seems to go to sleep if the mouse etc. isn't moving (while coming back and there's lots of VRAM being used and things going really slowly on the pc generally). Nothing more got output overnight a while after I left it even though I left the router on. Maybe it could be too high res. But it did quite a few at 896x576 just before I left it then soon after it stopped creating more. I don't know if it depends how many are in the queue.

koon grizzly

had this problem also, i need to be on the gradio app page or it doesn't update the preview window (but only in original framepack one click installer, i think framepack studio (so far) it always shows for me)

Juanmyth

Has this FP Studio any feature that is not present in your own FP gradio version?

patrikeon

Maybe stupid Q, can this be running locally without being online. Can’t get it working off. Thanks ☀️

SECourses

yes it has some extra features like queue but you can do queue with opening multiple gradio tabs - same url on different tab

patrikeon

Thanks for the answer 👍 hopefully solved, some day

patrikeon

Hi Furkan! Can I update to the latest, 0.4.1. Or maybe just minor changes, no speed or quality improvements? Thank you.

SECourses

i dont know sadly you need to try - this app not developed by me

Mordred Sillence

Hello. Is it possible to run two instances of framepack studio on multiple gpus on runpod?

SECourses

yes totally possible. when starting add this command export CUDA_VISIBLE_DEVICES=0 export CUDA_VISIBLE_DEVICES=1 first one will use first gpu and second one will use second gpu