Gemini 2.5 Pro LiveBench 屌打咗o1, claude 3.7 thinking

Everybodylies0

17 回覆

5 Like 1 Dislike

Everybodylies0 2025-03-27 00:37:49

power_off 2025-03-30 05:02:05

名師出鳩童 2025-03-30 12:38:24

大角倔喜 2025-03-31 08:04:03

Gemini都幾無聊，原本experimental改名flash thinking，跟住又改返名做experimental

Everybodylies0 2025-03-31 09:13:33

experimental 係non cot 喎一向
thinking 係reasoning
以後話唔再出non cot 先叫返做experimental

鄉廣史文 2025-03-31 09:27:12

GEMINI整合ANDROID
APPLE好玩完

sparda 2025-03-31 17:02:38

2.5 Pro OK喎
唔差過claude 3.7, 都快
而且免費, 用得過

正撚樣 2025-04-08 13:28:15

LLama4

sparda 2025-04-08 16:11:49

睇評價好垃圾喎
同qwen 32B比咋

Everybodylies0 2025-04-08 16:57:39

俾人屌到飛起

熱血軒達臣 2025-04-09 00:10:58

呢輪google贏番d口碑
Gemma 3 QAT 同gemini 2.5 pro 都唔差

sparda 2025-04-09 20:38:47

皮褸嘅乸媽3.1仲醒過4

正撚樣 2025-04-10 06:57:54

Meta 官網吹到自己好勁。

Everybodylies0 2025-04-11 20:19:09

好似又update 咗imagen 3
然後出咗veo 2
好快出flash 2.5
然後仲有chirp 3 同唔知咩

韋爵爺- 2025-04-11 23:51:26

Steve Jobs 死左之後蘋果一直都只係食老本

sparda 2025-04-18 13:13:18

flash 2.5出咗, 好平,但睇落唔醒

未必醒過o4-mini

熱血軒達臣 2025-04-18 18:13:33

弱少少平多多