Replit ModelFarm supports chat models, code models, and text embeddings. To find out more, check out our documentation or try out sample app. For now, Gen AI models are accessible for Pro and Hacker subscribers, so upgrade your account to get started. And, as always, comment below to let us know what you think!
No it is Palm2. It says that all the time. If it doesn’t know what to say,
40% of the time it gives a message about how it’s Palm2
40% of the time it says nothing
15% of the time it spews forth a bunch of gibberish and random characters intertwined with several unrelated sentences.
5% of the time it says a generic “I don’t understand” message Note that these are not official numbers but rather estimates based on my observations
Has documentation on the internal API been made yet?
Last time I checked; they were in v1beta.
Right after the release, another Replit Staff member said they were going to change the structure of the API a bit before they make Internal API documents, and would remind me, but I see none anywhere!!
Is there some place where I can find these Endpoints and structure a client without divulging into the official Modules?
Is this going anywhere? I understood ModelFarm was the flagship of Replit’s pivot to AI, so I expected something along the lines of OpenRouter, but native to Replit. But it’s what - Palm2 inference only?