While others said the platform would not work, it always worked for me. You have to get used the UI, but then it makes sense. You have to wait a little, as with other tools (for voices to be generated etc).
The tools splits your transcripts into small chunks, usually sentences. It then translates those chunks with AI, and the result is barely ok but not sounding good depending on the topic and source material. IT the spoken text sounds conversational — like any good business video should — it's harder to translate than an "article like, polished script". We don't speak like we write letters, that's the tricky part for translations. The cool thing is you can use your own prompts for revising the translation, and this makes it much better. It still needs a lot of work (as most good translations do).

Now for the voice part: Of course spoken Audio will have another duration in a foreign language: A good French translation and Spanish translation of the same English text will have different runtimes. This is the first thing where translate.video fails: Some fragments/chunks are too short, others are too long. So you speed up or slow down the audio, but it will sound artificial. You have to manually tweak this for each sentence. If it would use longer fragments, the trouble would even out: For example: Sentence 1 is shorter in the target language. Sentence 2 can be adjusted by a small tweak to be longer or shorter. And sentence 3 is longer. Currently, it's a mess. Using longer segments would even it out and all would sound good. Of course you can manually merge the segments but this is becoming way too much work.
It has features to make the speed differences less troublesome, but they just don't work well enough (OneTake.AI does a much better job with this, for example, it has an excellent voice clone and your Spanish, English and French videos will simply have different runtimes in the end, e.g. 3:10, 3:30 and 3:55 minutes, and by stretching the material it all looks good and very natural. Another alternative might be unmixr, which uses longer chunks and also let's you set an infliction and tones for the text to speech voices).

I could go on but there are two red flags: Carefully read the AppSumo offer, and the questions for premium voices and clones: It's ok if don't want to actually enable cloning for lifetime deals (they limit it to only 10 minutes even in higher plans, and each adjustment will again eat your full minutes of something needs to be edited). They just should pharse it that way and tell us the throughs. YOu can buy additional minutes for cloning for 1.5 USD per minute, which I again think is OK since the cloning soudns good and you do not have to pay this monthly, but only when you use it, and if you suddenly need 300 minutes today and then no minutes for the next 3 months this will work fine without restrictions. But they don't tell you even if you ask, you have to buy Tier 3 and look it up. Red flag.
Then they promised a much better translation they call "transliteration" or something, with more context/better auto-pre-translation-prompt. They give us a time frame - which they missed. It's ok, this happens. Just tell us. Then they did not answer a questions about the delay and when it will be available, they let it sit for a full week. Red flag 2.
Good bye and good luck. This could become better in the future, I provided detailed feedback so you know what you can improve (and for other buyers to see if you want it).