Tim Plant (@Timisscientist) profile picture
Joined November 2024
Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)Photo by Tim Plant (@Timisscientist)
Qwen 3.8-Max Preview 0721 Wow
176
Kimi has been too quiet for a while 🥲 @crystalsssup
1
3
971
Deepseek New imgae-identification mode, hmm..
995
Claude Opus 4.8 is another level
21
enemy of my enemy is my friend. @AnthropicAI @xai
23
Oh wow DS-V4 is cooler than CGPT-5.5
1
666
Poor Qwen3.6 max preview It claimed the knowledge cute-off date 2026-JAN However, it said:"GPT-5" not yet launched LOL
1
3
949
DeepSeek V4 will not only feature a trillion parameters and a million-token context window but also achieve deep adaptation to domestic chips such as Huawei's Ascend for the first time — marking a critical step for China's AI industry on the path to "decoupling from CUDA."
1
207
Breaking ✨ Congrats, Yu Bowen (prev: post-training head of Qwen), started his new journey in ByteDance (the company who trained seed model) All the best to Bowen. #qwen
2
90
My honest bottom line: GPT-5.4 is not “magical,” but it is elite-tier for research, coding, analysis, long-form drafting, and complex office-style work.
55
Tim Plant retweeted
GPT-5.4 Thinking and GPT-5.4 Pro are rolling out now in ChatGPT. GPT-5.4 is also now available in the API and Codex. GPT-5.4 brings our advances in reasoning, coding, and agentic workflows into one frontier model.
2,031
3,186
1,987
23,094
7,210,455
In recent months, Perplexity has made a series of decisions that feel hostile toward its user base. Whether you are a Pro member, a Max subscriber, or on an introductory trial, the experience has been consistently negative.
2
5
1,338
Thank you Junyang
24
Logan said more models are coming folks! Wen Gemini 3.0 GA / 3.2 ?
79
Quick test on Grok 4.20
2
5
1,327
Bro😣 Qwen3.5-plus Fail #qwen
2
3
1,027
Now, Deepseek named itself V3.3
1
8
1,562
Tim Plant retweeted
Minimax-M2.5 SWE-Bench Verified: 80.2% Multi-SWE-Bench: 51.3% BrowseComp: 76.3%
26
24
13
420
118,314
kimi 值得訂閱嗎?
0%Yes
0%No
0 votes • Final results
24