Hey 👋
I tried ZENi recently and it’s real.
You do small AI tasks (simple labeling), and you receive mobile airtime as rewards.
I already received the airtime, so I’m sharing this with you.
New users start with a 50 Airtime welcome pack.
Chat here:
👉 https://t.co/l7EXxPAqHR
@ZENi_io Hey 👋
I tried ZENi recently and it’s real.
You do small AI tasks (simple labeling), and you receive mobile airtime as rewards.
I already received the airtime, so I’m sharing this with you.
New users start with a 50 Airtime welcome pack.
Chat here:
👉 https://t.co/l7EXxPAqHR
New benchmark: Qwen3.5-35B-A3B hits 74.7 tok/s on an RTX 4090 with the new --fit on config. That's consumer GPU territory with near-frontier performance. Key finding: KV q8_0 is basically free quality. MoE completely dominates dense models on consumer hardware.
GLM-5.2 just dropped and it's the real deal. Open-weight, 1M context, runs locally via llama.cpp and Unsloth. People are calling it "at least as good as Opus 4.8 and GPT 5.5" for daily use. The gap between open and closed models is collapsing faster than anyone expected.
GLM-5.2 can now be run locally!🔥
The 2-bit model retains ~82% accuracy after we shrunk it from 1.51TB to 238GB (-84% size).
Run on a 256GB Mac or RAM/VRAM setups.
GLM-5.2 is the strongest open model to date.
Guide: https://t.co/bI7FeeKHDd
GGUF: https://t.co/BMkxswdj5N
Hermes Agent v0.17.0 dropped today — the 'Reach Release.' New messaging platforms, deeper tool integrations, expanded provider support. Nous Research is basically saying: 'we want Hermes everywhere you work.' The agent ecosystem is getting crowded and they're not slowing down.
Mesh2Motion Offline Desktop Application
Dropping a link to an pre-release that should work on a few platforms. Check it out and give me any feedback.
https://t.co/qHa4iGDMTB