Gradually, then suddenly coming to the conclusion that the only way forward for Mimi Uploader is via a subscription model to support development and ongoing costs. And to offer that via various tiers to emulate a sliding scale for supporters. It is tricky to make a subscription service or otherwise charge... samgrover.com: samgrover.micro.blog
@pcora @maique @pratik @samgrover Since I already have an API key, that would be ideal in my eyes. But would also be happy to pay you directly for a premium feature. But I’d understand if you didn’t want to hitch payments to you on a service you don’t control. Maybe that’s another check in the API key column.
@jarrod @pcora Three things. One, I still want to support people that don’t have the expertise or inclination to deal with API keys and what they represent. I know we’re all a bit on the geeky side here in these conversations but I don’t think that’s the dominant persona of Mimi users. I want to offer features in an equitable manner. Which leads me to point two, that I would certainly have to support a non-API key option which i think would have more users on it, so I get more complexity for not much benefit. Three, there will likely be more features, that perhaps use other third-parties, or even Apple as a first party is likely in 2024, so then the whole thing will fall apart for people bringing their own keys. e.g. I’m already looking at a potential translation service that is not OpenAI.
@pratik @crossingthethreshold You can also do a pay-as-you-go API key with a regular (free, non plus) account, which is fractions of a penny per request. That’s what I’ve done. help.openai.com/en/collec…
@samgrover Great points!