Yes, there is a very simple reason why I prefer to use Mistral AI over others and that is because it isn't moderated and allows me to use explicit language.Thanks for sharing your workflow. I take it that you're happier with Mistral's translation than say llama or chatgpt? I haven't looked at Mistral at all so am keen to hear about your experience.
ChatGPT will most likely not answer you and if you circumvent the content moderation too often you risk your account getting banned. Same applies for the OpenAI GPT API. You are required to use the content moderation API for Input and Output and if you send to many requests via API that have not gone through the content moderation API your account will also get banned.
LLAMA doesn't have the ban issue but it has the annoying "I don't want to talk about explicit stuff" attitude. There are custom models of LLAMA that have this "patched" out but in the end MIstral AI is as good as the newest LLAMA model.
Generally I recommend using something like OpenRouter (https://openrouter.ai/) where you can use most LLMs on the market and switch between them.
In the end I have found that Mistral Large v2 is currently the best LLM on the market that has no problems with explicit content and it has a large context window over 100k tokens, that is more than suitable for subtitle translations.