Sondo - AI Music Video & AI MV

You get realistic protagonist singing for 9 seconds or less. You can jiggle your shoulders or breathe or clench your fists but not much else. Then you will get the protagonist appearing in front of a scene. 4 seconds to 8 seconds. Not a whole lot of interaction. No way to change a wrong. System has a mind of its own. Even though I had a profile it gave me a random girl with a ponytail, the a random guy in shorts. After 4 attempts I got one I figured went sort of with the story. My advice is just live with it even if you have three hands or turn around and have a different face. Or have horns in every other scene. Even if you are hilariously driving the wrong way facing backwards as the scenery goes forward. You won’t get your way. CWA Computer Wins Again. Fun program to play with. Totally ignores my song. I tried over and over to get the song to show the ocean where the song said “ ocean”. Got it sort of close after 4 or 5 tries. That’s where all my credits go. Can’t have it your way. this isn’t Burger King. But I do have fun with it. I can see the time where I’ve gone through everything it can do and move on. My biggest peeve is nothing is 10 seconds long that I can use as a hook. One more second. Come on guys.

I purchased the yearly package with 8000 credits. Managed to make about 6 good music videos by simply adding a photo and a mp3 file. Noticed the program takes anywhere from 10 to 18 minutes to create a video. The video doesn’t really follow the story told by the lyrics. AI kept using the same or similar outfit. The lip singing voice is slightly off. In a few videos AI either added an extra hand or arm or object out of nothing. When placed inside a car, the scene or angle seemed backwards or something weird in the matrix. The developers have the concept right, but need to ensure things look normal and not from a science fiction movie. Overall, pictures and scenes look amazing.

If you’re considering Sondo as an AI video generation tool for anything beyond casual fun, it is not quite ready yet. The app has real potential, especially for turning music into quick videos, but it needs several more rounds of serious development before the current pricing feels justified. Lip-sync remains one of the biggest weaknesses. Characters’ mouths often fail to match the vocals properly. Sometimes the alignment looks decent, but more often the AI adds random or mismatched movements that pull attention away from the performance. On the visual side, the output frequently feels like a blend of two or three different generation models, with noticeable hallucinations, artifacts, and inconsistencies running through the entire video. It can be entertaining to experiment with, but it does not reach professional or even semi-professional quality. For pure entertainment, you can usually get better or more controllable results by creating individual clips or images with something like Grok’s video tools and then stitching them together yourself, often at a lower effective cost. The credit system also feels expensive for the output you receive. A single music video can consume around 1,500 credits, while lower-tier plans provide far fewer credits than that burn rate would suggest (for example, base plans around $10 for 800 credits, higher ones $30 for 3,000). If they reduced the entry price to about $9.99 per month and increased credits to something closer to 6,400, it would offer much stronger value. That kind of change could attract far more casual users, generate more revenue, and speed up the development of needed upgrades. Bottom line: Keep an eye on Sondo and check its progress in a year or two. It has the potential to become genuinely useful. For now, I would hold off on subscribing unless you are only dabbling for amusement and do not mind the current rough edges.