2892 位用户此时在线
24小时点击排行 Top 10:
- 本站自动实时分享网络热点
- 24小时实时更新
- 所有言论不代表本站态度
- 欢迎对信息踊跃评论评分
- 评分越高,信息越新,排列越靠前
5
Grok 4 Fast matches GPT-5 (High) performance at ~23x lower cost (95.7% cheaper)
2
1
1
Grok 4 Fast matches GPT-5 (High) performance at ~23x lower cost (95.7% cheaper)
Input tokens: 84% cheaper
Output tokens: 95% cheaper
Output speed: 150% faster
GPT-5: $1.25/M input, $10/M output
Grok 4 Fast: $0.20/M input, $0.50/M output
Grok: 344 tokens/s
GPT-5: ~138 tokens/s
btc
(
twitter.com)
•
X Freeze
6
Grok 4.1 Fast just made a nearly 3X jump in long-context accuracy
2
1
1
Grok 4.1 Fast just made a nearly 3X jump in long-context accuracy
A key challenge for AI models is that as chats get longer, most AIs forget. They lose the thread, make stuff up, or drop instructions.
Grok 4.1 Fast fixes this. And it’s a huge deal.
It was specifically trained
btc
(
twitter.com)
•
X Freeze
7
Today was the most productive coding day I’ve had in 10 years.
2
1
1
Today was the most productive coding day I’ve had in 10 years.
I spent 1.6k out of the 10k credits Cursor gave me, mainly using Opus 4.7 1M Max and GPT-5.5 Extra High Fast. The takeaway is simple. In Cursor these models feel unusually well tuned, fast, precise, and reliable for
IT技术
(
twitter.com)
•
Tw93
8
2
1
1
9
Chinese nationals spotted filming near Japan’s Ministry of Defense — reportedly attempting to “target” Self-Defense Force officers using facial tracking tools.
3
2
2
Chinese nationals spotted filming near Japan’s Ministry of Defense — reportedly attempting to “target” Self-Defense Force officers using facial tracking tools.
Imagine if this happened outside the Pentagon or the British MOD. Would any democracy tolerate that level of foreign
时政
(
twitter.com)
00:00:43