Subscribe
Sign in
Synthesia — Generate professional-looking AI videos from text in minutes
With Synthesia businesses can now create professional-looking videos at scale. Companies are using it for: 👨🏫️ corporate learning & communications, 👨💻️ personalized sales and customer success videos, 🛍️ e-commerce product videos, 🎓 e-learning and more.
150
Replies
Robin Good
For now I cannot get it to work. I have signed up for the $30 Personal account, but anytime I create a simple video and try to save it I get an undefined "something went wrong" error. I have already reported it and am looking yet to see whether I can get beyond this frustrating point.
Video: https://share.vidyard.com/watch/...?
Share
@robingood we'll be in touch! There was an overload on the system yesterday. So please try again this morning.
@steffen_tjerrild Thanks for the prompt follow-up. Just tried again to create a video, but situation looks identical to last night.
@steffen_tjerrild I have now done several tests and can tell you precisely what is causing that error. The commands for pauses like this one .
If I eliminate all of them, the audio voice over sucks but I can process and get a video.
P.S.: Do these test videos count toward my monthly quota if I delete them without downloading them?
@steffen_tjerrild @robingood Hey Robin, you found a bug in our system. This should be fixed within a few minutes now.
@steffen_tjerrild @jmarovt Yes! It has. Looks like things are working now. Will be testing more and will report.
P.S.: One thing not clear: how do you count the videos toward the quota. Is the count based on which ones I download or just on how many videos I create?
I ask this because I have created many test videos and would probably need to do more before I feel confident in producing one to be published. How does this work?
This is great. Is there a developer API? Would like to connect our audience to this.
This is pretty incredible technology that can significantly reduce costs of video production and localization.
Great launch!
This is super impressive! The audio doesn't sound robotic at all, and having the ability to translate the same generated video in multiple languages ca really help me expand our ads to new markets and get a foot in the door before we can launch there.
I found out about the company almost a year ago when they had a campaign with David Beckham, and it's impressive how fast they were able to package everything in a fully functional product...
Congrats to the whole team!
ps: are you working on different voice packages yet? Or is audio not your main focus right now?
@gidea Thanks Vlad. Yes, we are constantly adding more and better voices to the platform. Should improve very fast in the next months!
This is sooooo cool. Been really fun watching you shoot up the home page today and even playing with your product!
@vistikloft Yes, you can upload voice clips or use the text-to-speech voices. We plan to integrate AI voice cloning soon!
Super dope stuff. I have messaged my use case hope I get the access soon to try out. Can I know your charges?
Great stuff @vriparbelli @jmarovt, congrats on the launch! 💯
When is the API coming? :)
@vriparbelli @alencvisic Sooon.. https://www.synthesia.io/api We're starting to work with the first API customers in the coming weeks.
If it's this good now, I can only imagine how life-like it will be in just a few short years...congrats on your work to date!
Great team and tech. Synthetic Media and AI Video is delivering significant value to businesses around the world!
Wow, this seems epic. Especially I can see it being used for personalized videos at scale. Can you also use your own presenter - eg. people from our customer success team?
Synthesia is pretty gamechanging! We used an earlier beta version to turn one of our team into a programmable actor, then used it to a make a series of videos. Been waiting for Create so that this is all much easier - all you need is a script!
Just received the video I created as a marketing statement, but the expression of the statements made by the bot does not fit the context at all. The AI needs to understand the context lot better, needs more training.
I greeted Synthesia enthusiastically, as this is, a truly revolutionary technology that can help solve lots of situations while being highly cost-effective. After a few days as a paying customer, I still think the above, but I have realized I have been tricked into paying for a tech that it is NOT ready for prime time.
In synthesis, you cannot reliably use all of the voices being offered and on many of them you cannot use pause commands during the speech, which are 100% essential to make them sound realistic. Furthermore, certain voices available like Italian "original" sound totally artificial and have no resemblance to the quality of the samples being shown on the official site.
As soon as I signed up for a paid account I reported a critical bug that prevented saving any video when using pauses commands. I was publicly told, here on PH by Synthesia that I had found a bug and that it would be fixed within minutes.
Two days later I am back reporting that specific voices, including the one they had fixed the day before, are NOT working anymore if I use pause commands and I cannot save videos with the voices I want. To this, first I got no response for 24 hours from their support, then after soliciting again, they have responded the following:
"...this functionality currently works only with some of our voices.
They usually have "Natural", "Original" and "Default" in their name. Other voices will output an error if you try to add an SSML tag."
"It's unfortunate, but **we can't do anything about it right now as we use different underlying providers for voices**. And some support it and some don't. Over time more and more will support it though.
The first issue though - that you sent the video from. It's a similar issue with our voice providers, the best is simply to try later. We are doing what we can to pressure them on fixing their things."
Would you pay $30/mo for an unreliable service that does not match its landing page claims and that tells you all of the above privately and not before you pay for the service?
I am outraged.
To add insult to this mess I find it completely senseless that a tool which is not quite ready yet charges for the number of videos you produce independently of whether you download/use them or not. I don't see how can one test and train himself without doing this. $30/mo for not even been able to produce test videos looks to me as a bad proposition.
Finally, I think that the catalog of virtual actors, made exception for Nadia, really sucks. The other virtual actors do not look very professional at all, there is a lack of white people and I am just stuck with only character that would be appropriate for my audience type.