The prompts your GPT creates work very well. Congratulations and thank you so much! Before this GPT, I struggled a lot with my characters not walking, not moving and doing nothing, just getting frozen on the screen. Too bad this GPT arrives when I'm already almost finished my video! Just a small note, it should be instructed to produce prompts no longer than 2000 characters, of course we can ask ourselves to reduce them, it is just right for GPT to be complete. Thank you again.
This is incredible man!!!!! I used it on Minimax and it is giving me excellent results compared to when I prompt it myself. It gives me exactly what I want meaning I will take less time repeating good prompts. You are a GENIUS!!!!! It is people like you who make AI tools more exciting to use. You got a new subscriber in me.
another awesome video! Can't wait to watch more videos about this from you. Very few take the time or care with their videos and provide the value you do. Appreciate the custom GPT. Thanks.
LUV ALL YOU DO Ai BROTHA G! If I can suggest after using your new GPT the first 3 times, I would refine a touch so there is a space between 'Subject' parts as you can see here from it's last output for me: 'Scene Detail:An electrifying MMA arena at night, filled with a cheering crowd and vivid neon lights illuminating the space. The octagonal fighting cage is surrounded by spectators, with flashes of camera lights and glowing LED panels displaying dynamic graphics. The atmosphere is intense and vibrant, with a focus on the fighter's confident presence as he approaches the ring.Subject Details:A muscular UFC fighter with a commanding presence, wearing a striking neon blue and pink hooded jacket, paired with matching gloves and fight shorts. His beard is well-groomed, and he wears futuristic sports sunglasses, emphasizing his stylish and confident demeanor. His posture is upright, exuding self-assuredness as he walks toward the fighting ring.Style Description:A cinematic, high-energy style with sharp focus on the fighter and neon accents creating dramatic lighting effects. The colors are vivid and bold, with dynamic contrasts between the bright neon hues and the darker arena background. Keywords: cinematic intensity, neon lighting, bold color palette, dynamic sports atmosphere, modern MMA visuals.Camera Movement:A handheld-style camera tracks the fighter from a low angle, starting with a medium close-up of his upper body and then slowly pulling back to reveal his full confident stride as he walks down the aisle toward the cage. The focus remains on the fighter while the background softly blurs to en hance depth.Scene Motion:Flashing arena lights and LED graphics pulse rhythmically, while the crowd waves their hands and reacts energetically. Camera flashes flicker intermittently, adding to the excitement and energy.Subject Motion:The fighter walks with deliberate, confident strides, occasionally nodding to the cheering crowd. He adjusts his sunglasses briefly, smirking slightly as he approaches the cage, his posture radiating assurance that he will win the championship.'
First of all, thanks for the inspiring video - AGAIN! You‘re just the best! I was wondering if I could use the prompt generator and video tool to create different angles of my backgrounds. I‘m struggling with creating consistent backgrounds in midjourney because I have a complex setting at a playground… Even with your „expand method“ it does not work. Do you think it could work like this? 1) Use a midjourney image of the playground in the video tool 2) Create a video where the camera moves around 3) Use a still image from the video to continue in midjourney Thank you so much for your opinion and best wishes from Germany!
Good job 👍 on the video. I've used my COALS method forever which has worked well for me. Camera movement Object/Subject Action Location (when and where) Style
Yeah I like this breakdown. I just think it's missing a reminder/section to describe movement in the background. I do like a including reference to time period, but it may belong more as part of the style in my format over location. I'll think about it more
I tried to design the gpt based purely on clarity, reliability, and completeness. The intention was a good English description of video would be pretty model agnostic- but I couldn't test with everything. Is there a specific model where this format underperforming a different prompt style? I'd be interested in testing myself and trying to incorporate/improve the gpt...
@@glibatree definitely recommend playing with all the different frontier paid and open source mnodels if your intention is to make something thats more than just for sora. LLMs will almost always provide better prompts for video models than humans without "making" a gpt. I'd recommend setting it up to work with LTX video, and extremely popular, fast, n flexible open model, where a token length of 144 is excepted, and where an accurate description of the image input helps.
Editing artifact in the last copy, had to reupload. Sorry to anyone who was notified twice..
👋
Your art designer GPT has been such a huge help, so I'm very pumped to try this new one out. Thank you for all you do!
The prompts your GPT creates work very well. Congratulations and thank you so much! Before this GPT, I struggled a lot with my characters not walking, not moving and doing nothing, just getting frozen on the screen. Too bad this GPT arrives when I'm already almost finished my video!
Just a small note, it should be instructed to produce prompts no longer than 2000 characters, of course we can ask ourselves to reduce them, it is just right for GPT to be complete. Thank you again.
I've been using your GPT for quiet awhile now nice to see the video content.
This is incredible man!!!!! I used it on Minimax and it is giving me excellent results compared to when I prompt it myself. It gives me exactly what I want meaning I will take less time repeating good prompts. You are a GENIUS!!!!! It is people like you who make AI tools more exciting to use. You got a new subscriber in me.
Appreciate that!! Glad to hear it's doing well for you!!
another awesome video! Can't wait to watch more videos about this from you. Very few take the time or care with their videos and provide the value you do. Appreciate the custom GPT. Thanks.
No worries, it happens. All your videos are gold! Thank you. And thank you for the GPT!
Great video. Your GPT is fantastic! Thanks!!
You're the man!
Thanks for access to another great tool. :)
This is awesome, Waiting the New vídeo 🎉
Amazing. Thanks for the hard work you`ve must put in to create this. You deserve a Nobel Peace Prize :)
Top notch, you have really helped😊
This works really well. Testing it on Kling-only issue is the prompt generated is too large, so I have to shorten it.
I just love the random sparks on the rail after the boy skates off of it haha xD
Haha yeah, and the skater looks back for a second like, "Did I do that?"
The link for that CHATGPT doesn't work
Thanks for the video and GPT.
Great content. Thanks for sharing
new GPT model designed to generate high-quality AI video prompts. Interesting, thank you
LUV ALL YOU DO Ai BROTHA G! If I can suggest after using your new GPT the first 3 times, I would refine a touch so there is a space between 'Subject' parts as you can see here from it's last output for me: 'Scene Detail:An electrifying MMA arena at night, filled with a cheering crowd and vivid neon lights illuminating the space. The octagonal fighting cage is surrounded by spectators, with flashes of camera lights and glowing LED panels displaying dynamic graphics. The atmosphere is intense and vibrant, with a focus on the fighter's confident presence as he approaches the ring.Subject Details:A muscular UFC fighter with a commanding presence, wearing a striking neon blue and pink hooded jacket, paired with matching gloves and fight shorts. His beard is well-groomed, and he wears futuristic sports sunglasses, emphasizing his stylish and confident demeanor. His posture is upright, exuding self-assuredness as he walks toward the fighting ring.Style Description:A cinematic, high-energy style with sharp focus on the fighter and neon accents creating dramatic lighting effects. The colors are vivid and bold, with dynamic contrasts between the bright neon hues and the darker arena background. Keywords: cinematic intensity, neon lighting, bold color palette, dynamic sports atmosphere, modern MMA visuals.Camera Movement:A handheld-style camera tracks the fighter from a low angle, starting with a medium close-up of his upper body and then slowly pulling back to reveal his full confident stride as he walks down the aisle toward the cage. The focus remains on the fighter while the background softly blurs to en
hance depth.Scene Motion:Flashing arena lights and LED graphics pulse rhythmically, while the crowd waves their hands and reacts energetically. Camera flashes flicker intermittently, adding to the excitement and energy.Subject Motion:The fighter walks with deliberate, confident strides, occasionally nodding to the cheering crowd. He adjusts his sunglasses briefly, smirking slightly as he approaches the cage, his posture radiating assurance that he will win the championship.'
Humble A.I. Maestro TOP channel !!!! Merci
great vid, thanks.
First of all, thanks for the inspiring video - AGAIN! You‘re just the best!
I was wondering if I could use the prompt generator and video tool to create different angles of my backgrounds.
I‘m struggling with creating consistent backgrounds in midjourney because I have a complex setting at a playground…
Even with your „expand method“ it does not work.
Do you think it could work like this?
1) Use a midjourney image of the playground in the video tool
2) Create a video where the camera moves around
3) Use a still image from the video to continue in midjourney
Thank you so much for your opinion and best wishes from Germany!
Good job 👍 on the video. I've used my COALS method forever which has worked well for me.
Camera movement
Object/Subject
Action
Location (when and where)
Style
Yeah I like this breakdown. I just think it's missing a reminder/section to describe movement in the background.
I do like a including reference to time period, but it may belong more as part of the style in my format over location. I'll think about it more
@glibatree yeah, it depends. If you are using it for morning or night then I'd say style, but if you're using it for 300 BC then it's part of location
why i can t find th same GPT on the app on PC ? i only find it on the web version ?
thanks. you should mention straight up that these prompts are for sora. Most frontier video models have different prompt expectations
I tried to design the gpt based purely on clarity, reliability, and completeness. The intention was a good English description of video would be pretty model agnostic- but I couldn't test with everything.
Is there a specific model where this format underperforming a different prompt style? I'd be interested in testing myself and trying to incorporate/improve the gpt...
@@glibatree definitely recommend playing with all the different frontier paid and open source mnodels if your intention is to make something thats more than just for sora. LLMs will almost always provide better prompts for video models than humans without "making" a gpt.
I'd recommend setting it up to work with LTX video, and extremely popular, fast, n flexible open model, where a token length of 144 is excepted, and where an accurate description of the image input helps.
The web server reported a bad gateway error.
I❤YOU
This a great tool. But the AI video generator that you are using is not the best. And that’s being generous.
Which would you recommend? Kling?