Who has tried Mistral AI for article generation

TMONEX

Elite Member
Jr. VIP
Joined
Nov 9, 2013
Messages
1,758
Reaction score
764
I will like to add it to my AI course and just looking for validation if its worth the effort
 
I am using it regularly. Works great. Not as fast as OpenAI but no filters. Use the -large model for non-English content.
 
Not as great as OpenAI GPT-4, but has no censorship. Works smooth for us. I don't think it needs its own course as yet another text model though.
 
Not as great as OpenAI GPT-4, but has no censorship. Works smooth for us. I don't think it needs its own course as yet another text model though.
No in my course, you can add several models to one setup. So I want to add Mistral. in another chapter and how to connect it to their current setup.

I am not sure of the advantages of censored and uncensored contents since I am not in adult niche.

Thanks for your input
 
Since many autoblogging tools today work via OpenRouter, they allow one to add about 100 various AI models in one setup. Just change the model ID and it will work using the same API: https://openrouter.ai/docs#models

Most importantly, all of these models are managed through a single payment account. Could it be any simpler?

OpenRouter rocks!
 
Thanks for the info. If it doesnt have censorship its worth a try. I find my self having trouble creating product reviews of adult products simply cuz the ai refuses to talk about it.
 
I am using it regularly. Works great. Not as fast as OpenAI but no filters. Use the -large model for non-English content.
Great for adult niche and can generate ton of content for gsa ,rankerx etc with scm....scm uses their API..but quality wise open ai hits ..
 
Seems video didn't give the conclusion?

It was enough for me to come to my own conclusion. The texts generated by Claude 2 for the same prompt with the same temperature and max token limit are not as good as the ones made by OpenAI GPT-4 Turbo. On the other hand, it doesn't explore Claude 3, specialty the Opus model, so maybe it really has been improved since then...

So feel free to do your own tests in the OpenRouter Playground: https://openrouter.ai/playground
 
Back
Top