I still have some GEN-3 credits available! If you have some ideas for prompts of what you'd like to render just leave them within the comments below + up vote the ones you'd like to have me generate, i'll circle back in a couple days and generate the best ones :)
I did this yesterday and also bought another $10 in credits. I used almost the same prompt for every video. I was trying to get a treehouse with the camera dollying back. I prompted 'treehouse' 3 times in most prompts yet only got one in about 1/5 of the gens. Also prompted low-angle camera and half of the gens were mid to bird's eye. Generally the videos are really good but what I'm seeing others say is that it's just not following prompts very well. I get that it's alpha but on the flip side, we are paying real money so it should be a little better in this regard. Gen 4 should be pretty amazing but I don't think 3 is quite there.
I have been working with Runway a lot. I'd like to see what you can do with consistency between scenes. Even using the same prompts and seed, not having good results myself.
I have purchased and cancelled the subscription at the same day means today when will I get the refund? did any of you cancelled your subscription and got the refund. or i need to do something more to get the refund.
Thank you for this video. I think for the amount, there is not enough for credits for generation, especially if the tool isn't giving you what you want. I'm assuming that if you click generate and it doesn't give you what you want, then it still takes you credits? Or is it that once you accept and download you're charged the credits?
They have higher tiers where you can get more generations, you just have to spend more money. I'd only suggest that if you're planning on being a power user and you want to make content around AI videos/films. I also think image-to-video is better because it offers you more control over the end result, so you're using fewer tokens on re-runs. So maybe the move is to generate an image in Midjourney of a camera on a butterfly at the angle you like, then pass that image to Runway to turn it into a video.
I am curious if Runway Gen 3 can create a character. I was looking to create a character with a realistic face and use it in many UA-cam videos. Not sure if Runway can do this yet.
They released a plan with unlimited generations of Gen-1, Gen-2 and Gen-3 Alpha in Explore Mode at relaxed rate at 95$ per month, or $76 yearly. I feel it's the good plan if you want to experiment as much as you want.
Whenever you try to generate a video, use short and concise prompts - it will result in a more random video, but at least is something that usually looks much better in comparison to when providing detailed prompts; I had the same problem using luma…and to be honest, I still find luma better…it could be just a personal opinion, but I hope that in the following months video generation will be much better
@@DevelopersDigest here is a video that I uploaded ua-cam.com/users/shorts8ydVfNvGEWU?si=WHH_GBuw73SwhxoT where I point out explicitly what was used in order to generate the videos: prompt or image; it still does some crazy morphing and the hands usually look extremely weird, but hey - when you think what we had at our disposal 1 year ago, and what we have now - is a great advancement in this field
@@DevelopersDigest ps. I forgot to mention the following: if you are going to try it out, DON’T choose “Prompt enhancement” button when generating the video
Thats ridiculous expensive. Once the open source alternatives will be available, I will may be try. But I will not even look at this. I make my videos my self in much higher quality at the moment..don't need such shit
@DONTPANICx yea. That should've been mentioned in the video. The tier shown in the video is the lowest tier, there are other tiers that cost more money for more generations.
The biggest problem with Gen-3 right now is prompt following. It doesn't take directions well, which really counts when the prompt limit is 500 characters and there's no image to video modality. I spent all my Runway credits on Gen-3 videos this month and exactly none of them were usable! Cute toy, not worth it.
Add picture and then types long prompt. Gets weirdly confused and faced with zoom out and nothing happens. Not sure.What the hell i'm doing wrong with this damn prompt
It's too expensive and the quality is hit or miss. When it hits, it's magical, but the misses are always so disappointing (like, it'll just have ignored parts of what i told it). If Runway can make it cheaper, then I'm fine to keep plugging away hoping to get a happy accident. Until then, its overpriced.
@@Kreative_Katz Case and point infancy. I’m sure this time next year we will be able to make top quality professional work. But as of right now we’re still in the sandbox, playing with finger paint and crayons.
It depends if these systems don’t only get better with high quality video output, but if they will be actually able to closely adhere to prompts, especially to those that require details and uncommon motives. Cause right now I only see a production of vague, fairly unrealizable, yet better looking generic content that might or might not be somewhat close to what was described to the system
When are they going to open up the image to video feature? without image to video its kind of useless.... you can let the model create all sort of beautiful stuff, but it's not what you NEED. if I could take a photo and bring it to life, that's make it a tool. Luma can make that, but the rendering times there are INSANE.... could reach 14 hours for a 30 sec video
a BIG warning to users. the credit system IS pricey. if you go by credits alone for gen3 you WILL end up spending a LOT to get what you really want. THE BEST POSSIBLE OPTION BY FAR is to get unlimited. its slower but you can make as many videos as you want and have all the time to fine tune it.
I'm considering the unlimited, any idea how long does it takes to generate 10s in relaxed mode? Also, how many times do you have to re-prompt to get a usable video (on average) Thanks :)
@@aiStoleMyJob it varies through the day. Late at night you can run 4-5 concurrent ones. It does sometimes take a few tries to dial in desired looks, which is why I recommend unlimited to anyone seeking to use it. All other plans will leave people feeling shorted and like it's a waste IMO
I still have some GEN-3 credits available! If you have some ideas for prompts of what you'd like to render just leave them within the comments below + up vote the ones you'd like to have me generate, i'll circle back in a couple days and generate the best ones :)
do an olive seed growing and then time lapsing the growth of the olive tree from seedling to full grown olive tree
It would be nice if it would give you a low res preview so you can approve it before generating the full thing
this way they make more money
I did this yesterday and also bought another $10 in credits. I used almost the same prompt for every video. I was trying to get a treehouse with the camera dollying back. I prompted 'treehouse' 3 times in most prompts yet only got one in about 1/5 of the gens. Also prompted low-angle camera and half of the gens were mid to bird's eye. Generally the videos are really good but what I'm seeing others say is that it's just not following prompts very well. I get that it's alpha but on the flip side, we are paying real money so it should be a little better in this regard. Gen 4 should be pretty amazing but I don't think 3 is quite there.
I have been working with Runway a lot. I'd like to see what you can do with consistency between scenes. Even using the same prompts and seed, not having good results myself.
It's definitely not worth
But i think we will get there when it's much cheaper and the results are a lot better
I have purchased and cancelled the subscription at the same day means today when will I get the refund? did any of you cancelled your subscription and got the refund. or i need to do something more to get the refund.
Thank you for this video. I think for the amount, there is not enough for credits for generation, especially if the tool isn't giving you what you want. I'm assuming that if you click generate and it doesn't give you what you want, then it still takes you credits? Or is it that once you accept and download you're charged the credits?
Not there yet. But it’s getting super close
Beg to differ. Very far from acceptable to professional filmmakers at present. But our hopes remain high.
@@davidskybrody agreed. Started trying to use It in a documentary and It wasn’t useable 😢
Apparently Gen3 is like a 3 year old toddler when it comes to following directions 😂😂😂
They have higher tiers where you can get more generations, you just have to spend more money. I'd only suggest that if you're planning on being a power user and you want to make content around AI videos/films. I also think image-to-video is better because it offers you more control over the end result, so you're using fewer tokens on re-runs. So maybe the move is to generate an image in Midjourney of a camera on a butterfly at the angle you like, then pass that image to Runway to turn it into a video.
I am curious if Runway Gen 3 can create a character. I was looking to create a character with a realistic face and use it in many UA-cam videos. Not sure if Runway can do this yet.
They released a plan with unlimited generations of Gen-1, Gen-2 and Gen-3 Alpha in Explore Mode at relaxed rate at 95$ per month, or $76 yearly. I feel it's the good plan if you want to experiment as much as you want.
thank you - the price 15$ for 6 videos is very high - but if I do not try I will not have my own opinion about it:)
15 is a scam. go for 95$ subscription. it's unlimited
...and the 6 videos generated are not assured of the aesthetic quality! It's like an expensive casino slot machine.
definitely not worth it but really amazing what this is capable of! maybe in the future it will become more affordable
Whenever you try to generate a video, use short and concise prompts - it will result in a more random video, but at least is something that usually looks much better in comparison to when providing detailed prompts; I had the same problem using luma…and to be honest, I still find luma better…it could be just a personal opinion, but I hope that in the following months video generation will be much better
Thanks for the tip - I haven't tried Luma. What do you like about Luma in particular?
@@DevelopersDigest here is a video that I uploaded ua-cam.com/users/shorts8ydVfNvGEWU?si=WHH_GBuw73SwhxoT where I point out explicitly what was used in order to generate the videos: prompt or image; it still does some crazy morphing and the hands usually look extremely weird, but hey - when you think what we had at our disposal 1 year ago, and what we have now - is a great advancement in this field
@@DevelopersDigest ps. I forgot to mention the following: if you are going to try it out, DON’T choose “Prompt enhancement” button when generating the video
Does it create imaginary or phantastical things well? Or just stock video?
So we can generate only one video of 62 secs?? Only one video in entire month
It’s a credit based system. You are able to choose between 5 second and 10 second generations 🙂
@@DevelopersDigest So how many seconds can we generate for a 15 dollars package in a month
Thats ridiculous expensive. Once the open source alternatives will be available, I will may be try. But I will not even look at this. I make my videos my self in much higher quality at the moment..don't need such shit
@DONTPANICx yea. That should've been mentioned in the video. The tier shown in the video is the lowest tier, there are other tiers that cost more money for more generations.
The biggest problem with Gen-3 right now is prompt following. It doesn't take directions well, which really counts when the prompt limit is 500 characters and there's no image to video modality. I spent all my Runway credits on Gen-3 videos this month and exactly none of them were usable! Cute toy, not worth it.
ill buy the unlimited as soon as it has extend
Good demo. Why it even not followed the exact wordings? 5:02 😅
Add picture and then types long prompt. Gets weirdly confused and faced with zoom out and nothing happens. Not sure.What the hell i'm doing wrong with this damn prompt
You should have used quotes for 'Developers Digest'
I'll try again with that and see if that helps!
SMD, Gen 3...I'll stick with Luma.
It's too expensive and the quality is hit or miss. When it hits, it's magical, but the misses are always so disappointing (like, it'll just have ignored parts of what i told it). If Runway can make it cheaper, then I'm fine to keep plugging away hoping to get a happy accident. Until then, its overpriced.
Loving Runway so far!
So for I’m not that impressed with Gen-3 It’s obvious this technology is STILL in its infancy stages, maybe in a few more months.😊
Yeah, but to be fair it's Gen 3 alpha. They're nowhere near finished with it.
@@Kreative_Katz Case and point infancy. I’m sure this time next year we will be able to make top quality professional work. But as of right now we’re still in the sandbox, playing with finger paint and crayons.
@@Kemetrical agreed.
It depends if these systems don’t only get better with high quality video output, but if they will be actually able to closely adhere to prompts, especially to those that require details and uncommon motives. Cause right now I only see a production of vague, fairly unrealizable, yet better looking generic content that might or might not be somewhat close to what was described to the system
I should have watched this video before purchasing. :/
Yea its def WORTH, for the speed alone easy. But if youre picky then prob not worth
nice vid
Thanks!
I tried Gen3 too...personally, not worth it:
Gen3: 15$ month ,=12 generations.
Luma: 10$ =70 !!!
Gen3 is probably.... ¿microscopically better? NO contest, IMO
When are they going to open up the image to video feature? without image to video its kind of useless.... you can let the model create all sort of beautiful stuff, but it's not what you NEED. if I could take a photo and bring it to life, that's make it a tool. Luma can make that, but the rendering times there are INSANE.... could reach 14 hours for a 30 sec video
It wasn't as good as I expected aye, the demos were very cherry picked.
a BIG warning to users. the credit system IS pricey. if you go by credits alone for gen3 you WILL end up spending a LOT to get what you really want. THE BEST POSSIBLE OPTION BY FAR is to get unlimited. its slower but you can make as many videos as you want and have all the time to fine tune it.
I'm considering the unlimited, any idea how long does it takes to generate 10s in relaxed mode?
Also, how many times do you have to re-prompt to get a usable video (on average)
Thanks :)
@@aiStoleMyJobfrom 30 sec to 2 min for 1 vid... and it'll give you 100% whatever you ask depending on your prompt or 0% ...
@@aiStoleMyJobfrom 30 sec to 2 min for 1 vid... and it'll give you 100% whatever you ask depending on your prompt or 0% ...
high price for this wonky video result.
only unlimited is worth it. ive done almost 1.5k generations since gen3 release LOL
That's amazing, how long does it typically take to generate 10s in relaxed mode?
@@aiStoleMyJob 2-3 minutes if that. Mostly faster
@@iamisobe That's pretty good. I think you can have 2 running at the same time right? How long does it take for you to get something usable.
@@aiStoleMyJob it varies through the day. Late at night you can run 4-5 concurrent ones. It does sometimes take a few tries to dial in desired looks, which is why I recommend unlimited to anyone seeking to use it. All other plans will leave people feeling shorted and like it's a waste IMO
@@iamisobe How many generations can you have running at the same time?
Not worth it
Nah I tried it too but it wasn’t worth it.
Maybe the price is high but if you have the time and the knowledge you can do magnificent things
And the money 😂😂😂
Your prompts were pretty bad, so this review isn't that informative.
Sorry!
Not worth it !!
NOT WORTH IT 🖤 Real SCAM
I paid before watching this. I’m not impressed
Not worth it