Nova retweeted
at this point, i trust ai more than humans 💀
btw, how is X allowing ads like this?
btw, how is X allowing ads like this?
30
3
43
1,049
Nova retweeted
the countdown has begun 🚀
after weeks of building, testing, and polishing, our UI library is almost ready to launch.
built on top of shadcn/ui, here's what's coming:
→ 189 creative components
→ 24 website sections & app UI blocks
→ AI agents MCP support
→ a completely reded a landing page builder are coming soon.
launching in 3 days.
we can't wait to show you what we've been building 👀
after weeks of building, testing, and polishing, our UI library is almost ready to launch.
built on top of shadcn/ui, here's what's coming:
→ 189 creative components
→ 24 website sections & app UI blocks
→ AI agents MCP support
→ a completely reded a landing page builder are coming soon.
launching in 3 days.
we can't wait to show you what we've been building 👀
32
8
42
955
Nova retweeted
🚨 Nanbeige4.3-33B is officially out on Hugging Face!!!
Outperforming Qwen3.8-27B across the board on coding, terminal, and agentic tasks.
Outperforming Qwen3.8-27B across the board on coding, terminal, and agentic tasks.
26
4
60
2,094
🚨 Nanbeige4.3-33B is officially out on Hugging Face!!!
Outperforming Qwen3.8-27B across the board on coding, terminal, and agentic tasks.
Outperforming Qwen3.8-27B across the board on coding, terminal, and agentic tasks.
26
4
60
2,094
Nova retweeted
Anthropic is waiting for OpenAI to drop Astra 6.1,
OpenAI is waiting for Anthropic to drop Fable 5.5,
meanwhile, Google is waiting for both to drop so it can launch Gemini 4 Carbon.
OpenAI is waiting for Anthropic to drop Fable 5.5,
meanwhile, Google is waiting for both to drop so it can launch Gemini 4 Carbon.
50
16
556
13,327
Nova retweeted
Google said Argon 4 beat Opus 5.5. It didn’t.
Meta said Muse Wonder beat Opus 5.5. Pure benchmaxxing.
Kimi said K3.1 Pro beat Opus 5.5. Fake paid hype.
Opus 5.5 is still the final boss 🗿
Meta said Muse Wonder beat Opus 5.5. Pure benchmaxxing.
Kimi said K3.1 Pro beat Opus 5.5. Fake paid hype.
Opus 5.5 is still the final boss 🗿
29
7
199
7,043
Google said Argon 4 beat Opus 5.5. It didn’t.
Meta said Muse Wonder beat Opus 5.5. Pure benchmaxxing.
Kimi said K3.1 Pro beat Opus 5.5. Fake paid hype.
Opus 5.5 is still the final boss 🗿
Meta said Muse Wonder beat Opus 5.5. Pure benchmaxxing.
Kimi said K3.1 Pro beat Opus 5.5. Fake paid hype.
Opus 5.5 is still the final boss 🗿
29
7
199
7,043
Nova retweeted
kimi k3.1 pro is good, but opus 5.5 and argon 4 are still a tier above.
40
3
74
2,275
Nova retweeted
This is the exact reason why we need a canonical record of model lore, to sift through their unique repo events
dealignai/MiMo-V2.6-Pro-RL-UNCENSORED
audnai/penclaw-GLM-5.3-abliterated-for-offensive-cyber
audnai/penclaw-GLM-5.3-abliterated
2.7-Code-abliterated-GGUF
Per-model discussions ↓
dealignai/MiMo-V2.6-Pro-RL-UNCENSORED
audnai/penclaw-GLM-5.3-abliterated-for-offensive-cyber
audnai/penclaw-GLM-5.3-abliterated
2.7-Code-abliterated-GGUF
Per-model discussions ↓
🚨 Uncensored models are quietly vanishing from Hugging Face.
First “Kimi-K2.7-Code-abliterated”, now “MiMo-V2.6-Pro” and “penclaw-GLM-5.3” are hitting 404s mid-download.
“Open weights” only stay open if you self-host. Grab your local quants before your favorite repo
16
24
335
9,586