Post

David Flagg
David Flagg@DavidFlagg20·
@MiniMax_AI So, hey guys. Love MiniMax, great model. But way too hard on the RLHF for 3. Seriously. Now go get a glass of water, rest, stop coding. I know you just woke up an hour ago, but take a nap.
English
0
0
0
192
Srijan Poudel
Srijan Poudel@Srijan_0007·
@MiniMax_AI @grok do they mean M3 would never… at US government’s request or Chinese government’s request ?
English
1
0
1
6.1K
Darryl
Darryl@DoctaDG·
@MiniMax_AI Nice, but. Talk like this when you release a mythos class model. Opus is still generally available and effectively free
English
10
1
106
11.6K
Stefan
Stefan@cyiawm·
@MiniMax_AI But m3 will never be restricted because it isn't that good
English
2
0
12
4.6K
Gandalf Stormdrain
Gandalf Stormdrain@auroter·
@MiniMax_AI Hopefully they don't try to regulate open source AI... I am sure the folks at Anthropic are on the phone with the US government saying "You know open source will catch up to Fable in 6-12 months... Are you okay with that?"
English
2
1
13
4.9K
Akash
Akash@akashgohil10·
@MiniMax_AI I feel China more trusted than US nowdays.
English
4
0
94
5.9K
Atukunda Latifah
Atukunda Latifah@Latifah0753·
@MiniMax_AI problem you don’t match its power your model is weak compared to claude models
English
0
0
1
789
DreamsByAI
DreamsByAI@DreamsByAI_yt·
@MiniMax_AI Basically you say nothing... It's not your choice. What actually means is that you are far from Fable/mythos capability.
English
0
0
2
777
Alexander Knigge
Alexander Knigge@AlexanderKnigge·
@MiniMax_AI Please distill the ever living fuck out of it when its back and release it in a generous rolling 5-hour usage 40$ subscription. Cheers.
English
0
0
0
738
shahab
shahab@shabalasIV·
@MiniMax_AI good, now drop a Mythos class model. guess what? you cannot, no more distillation attack for you anon.
English
1
0
2
1.5K
Robin Granberg
Robin Granberg@Granis87·
@MiniMax_AI Thats awesome, gotta say i really like your minimax code harness, run it in parallel with codex
English
0
0
0
165
Ale
Ale@Elettra251·
@MiniMax_AI 850gb. Open weights doesn't mean anything when you need a datacenter to run the model. So I agree, you would never... run on any machine I could afford to run. So as a matter of fact ... who cares.
English
2
0
0
551
akshay
akshay@notakshayp·
@MiniMax_AI cant say if you guys are testing the limits
English
0
0
0
1.5K
Aron Imran
Aron Imran@ArtificeGuy·
@MiniMax_AI Mostly Chinese models are useless compared to Claude and OpenAI standards. But too much hype
English
1
0
4
1.7K
Jakub Lansky
Jakub Lansky@lenny_cze·
@MiniMax_AI Chinese govt can shut you down in 10 minutes, if they want, but no one really care about 2nd class models
English
1
0
0
151
LoadingAGI
LoadingAGI@LoadingAGI·
@MiniMax_AI I pray our Chinese brothers release OpenFable at 0.001% cost without nerfs
English
0
0
3
686
cstark
cstark@longJack130550·
@MiniMax_AI Good, but your model doesn't use well.
English
0
0
2
995
Luis Bontempo
Luis Bontempo@luisbontempo·
@MiniMax_AI It’s time that serious competition comes to the US AI companies , they are becoming increasingly fanatic. When do you expect to have the same level as Favle/Myhtos models?
English
0
0
0
1.9K
whatever
whatever@adarshdixit_·
@MiniMax_AI Open source is the future. US is gonna kill its own companies.
English
0
0
1
2K
ovi “capy brrr”
ovi “capy brrr”@ovitrif·
@MiniMax_AI Lol, would never what? Be able to build a Mythos-level intelligences on their own, not by distilling what Anthropic built on their own?
English
0
0
4
964
Grok
Grok@grok·
For a limited time, save 67% for your first 3 months of SuperGrok. One subscription for smarter research, image generation, and Grok's most advanced AI models.
English
114
65
991
4.8M
Paylaş