English
NewOneZ
7 posts
关于Qwen @Alibaba_Qwen 开源的最新消息：
尽管大家都很想要Qwen3.7的开源版本，但据我所知，目前没有任何决定。
自从3月份的人事变动以来，新的管理层并没有决定是否继续坚持开源Qwen文本模型（Qwen3.7和后续的模型）。
昨天我看到很多人在转发Qwen研究员 @xuanmingzhangai 的截图，但他也承认这个个人观点。
也许社区的声音（压力）会让新的领导层尽快完成思考。
希望Qwen还可以继续开源，你们到底知不知道社区里有多少项目都依赖于Qwen系列啊。。。
xuanming zhang@xuanmingzhangai
@AiBattle_ Well, it’s still a personal view, but thanks for all the support!
中文
@MrSubstrate @Kaladivo @michal_novak_21 I use AI for fun, don't care about distillation. Chinese models were trash compared to GPT 5.2/Opus 4.5 at that time. Now they literally match Opus 4.5 level. Qwen 3.7 performs good most of time. Ironically, Opus 4.6 was nerfed into a Chinese model.
English
@NewOneZ_ @Kaladivo @michal_novak_21 Have you actually used them?
They're very Chinese: just distilled from GPT and Claude. It's a classic Chinese knockoff scenario, but with intelligence
English
Vláda USA zakázala Fable 5, protože je "příliš inteligentní" a před 2 dny se ho podařilo jailbreaknout.
Geniální tah, opravdu. Sebrat běžným lidem přístup k inteligenci a doufat, že čínský open-source, který zaostává jen pár měsíců, Mythos rodinu nikdy nedožene...
Dožene, a bez jakýchkoli zábran. Nic nevyřešili, jen oddálili.
Čeština
@Kaladivo @michal_novak_21 The gap between Chinese and US models is only about six months. Sure, Mythos is way better than anything else out there, but a Chinese model at the Opus 4.6 level is still production-grade. You can't call Opus 4.6 trash the moment Mythos drops.
English
@michal_novak_21 Bohužel opensource modely (nevím, jak ty čínské closed source), jsou dost daleko od opensource modelů. Například tento benchmark do hezky demonstruje: deepswe.datacurve.ai
Čeština
Muse Spark at #3, better than GPT-5.5?
I’ve never seen a single person using Muse Spark.
Nobody takes Arena seriously anymore.
Arena.ai@arena
The top 5 labs in Text Arena rankings by category show that frontier models have distinct strengths and tradeoffs. #1 @AnthropicAI, Claude Opus 4.7 - The most consistently dominant model overall, leading top-tier across nearly every major category. #2 @GoogleDeepMind, Gemini 3.1 Pro - Well-rounded, with a notable edge in Creative Writing, ranked below Opus 4.7 and GPT-5.5 High in Expert #3 @AIatMeta, Muse Spark - Particularly strong in Overall and Coding, though it’s lagging behind in Expert tasks, Math, and Longer Query performance. #4 @OpenAI, GPT-5.5 High - One of the most balanced models overall, staying competitive with the top two across most categories, with especially strong performance in Expert and Math. #5 @xAI, Grok 4.20 - A more specialized profile, standing out primarily in Creative Writing and Hard Prompts, while lagging behind in Expert tasks.
English
Wan 2.7 is coming soon! A major all-around upgrade over 2.6! 🙀
What can we expect?👀👇🏻
- first-frame & last-frame video generation
- 9-grid image-to-video
- subject + voice reference
- instruction-based video editing
- video recreation / replication
There are still two weeks left in March, and there's still so much possibility! 😸
Alisa Qian@alisaqqt
Wan 2.7 is planned to launch within March — and it’s a major all-around upgrade over 2.6. Wan 2.7 will support: - first-frame & last-frame video generation - 9-grid image-to-video - subject + voice reference - instruction-based video editing - video recreation / replication A more powerful and comprehensive creative workflow is on the way.
English