Basic @rlybasic

Joined November 2022

Tweets

18
Followers

7
Following

138
Likes

12K

Basic @rlybasic

3 days ago

@Repvbblica @yandex, почините уже человеку лампочку!

0 0 0 25 0

View Details

Basic @rlybasic

5 days ago

@AteoBreaking В Казахстане такое давно практикуется 🤷‍♂️

0 0 0 222 0

View Details

Basic @rlybasic

7 days ago

@midnight_bat86

0 0 3 584 0

View Details

Basic @rlybasic

2 weeks ago

@punkratiev

0 0 6 97 0

View Details

Basic @rlybasic

2 weeks ago

@robiartec @VadimStrizheus Nah, I think it’s between opus-4.8 and 4.7. I’m comparing the results of working with versions 4.8 and kimi-2.7 on my own projects, and the difference is simply enormous.

0 0 0 44 0

View Details

Basic @rlybasic

2 weeks ago

@jun_song I think the same 💯

0 0 0 62 0

View Details

Well, seriously, I’ve been using Claude on the max plan for a year now, and it’s really gone downhill. I don’t know what benchmarks they run there, but it’s been going completely haywire lately – one minute it’s making up some context, the next it’s drastically changing the formatting style of methods, and then it starts renaming arguments. I switched to Kimi a week ago, and there are no more misunderstandings; the model really does work better, and nobody’s degrading it every week.

0 0 0 291 0

View Details

Basic @rlybasic

3 weeks ago

@Zai_org Are any benchmarks with difference between Kimi-2.7?

0 0 0 408 0

View Details

Basic @rlybasic

3 weeks ago

@jumperz Where’s kimi-2.7 😭

1 0 0 2K 0

View Details

Basic @rlybasic

3 weeks ago

@HonestDadonX @ClaudeDevs

1 0 1 96 0

View Details

Basic @rlybasic

3 weeks ago

@ogivus Шляпа полная. Если ты в сфере инфобеза, то тебя постоянно кидает на опус 4.8, который каждый запрос расценивает как «не легитимный» и ломает сессию.

0 0 8 1K 0

View Details

Basic @rlybasic

3 weeks ago

I work in the field of information security, and since the integration of Anthropic’s CVP programme, it has become virtually impossible to use; almost every query I make is flagged by the security system. The only model that works properly is Sonnet 4.6, because it doesn’t have a filter. I’m already considering switching away from Claude