0.5T total. Current Grok is half the size of Sonnet and 1/10th the size of Opus. Very strong model for its size.
GENERATIVE AI
-
Nous Research CTO Space Recording Discussion
By
–
Yes! You should listen to the recording of the space we did earlier with @NousResearch
’s CTO. -
Free ML AI Resources Available for Subscribers Monthly
By
–
I continue adding more 𝐅𝐑𝐄𝐄 resources for my @X Subscribers to access: 𝗼𝘃𝗲𝗿 𝟰𝟴𝟬𝟬 𝗿𝗲𝘀𝗼𝘂𝗿𝗰𝗲𝘀 𝗮𝗿𝗲 𝗻𝗼𝘄 𝗮𝘃𝗮𝗶𝗹𝗮𝗯𝗹𝗲. Subscribe for only US$6/month #MachineLearning #ML #AI #DataScience #Python #Mathematics #Statistics #SQL #GenAI #AgenticAI
-
Capability Overhang: Models vs Realized Value for Developers
By
–
Capability overhang is the delta between what models can do vs. the value users realize
— Amol Avasare (@TheAmolAvasare) 9 avril 2026
People often see this as only an end-user problem, but it applies just as much to devs
e.g. when building agents and harnesses for long-running tasks, the supporting infra is typically quite… https://t.co/w6IiiIEwn3Capability overhang is the delta between what models can do vs. the value users realize People often see this as only an end-user problem, but it applies just as much to devs e.g. when building agents and harnesses for long-running tasks, the supporting infra is typically quite
-
GPT-5.4 Hermes Agent Integration: Zero Configuration Setup
By
–
GPT-5.4 in Hermes Agent is surprisingly GOOD.
— Shubham Saboo (@Saboo_Shubham_) 9 avril 2026
You can even import your OpenClaw settings, skills, memories and API keys in seconds.
Zero extra config. It just works. pic.twitter.com/eO65gitYfLGPT-5.4 in Hermes Agent is surprisingly GOOD. You can even import your OpenClaw settings, skills, memories and API keys in seconds. Zero extra config. It just works.
-
GLM 5.1 Leads Open Source Model Leaderboard
By
–
GLM 5.1 TOPS THE OPEN SOURCE LEADERBOARD 🚀🚀
— Bindu Reddy (@bindureddy) 9 avril 2026
GLM 5.1 is the new open source leader and is comparable to closed source models like Opus and GPT 5.4
Open source is catching up to closed source every day and GLM 5.1 performs really well on both the coding and the agentic… pic.twitter.com/9TSKSNs8VpGLM 5.1 TOPS THE OPEN SOURCE LEADERBOARD 🚀🚀 GLM 5.1 is the new open source leader and is comparable to closed source models like Opus and GPT 5.4 Open source is catching up to closed source every day and GLM 5.1 performs really well on both the coding and the agentic benchmark An excellent option for coding agents and open claw!
-
US labs abandon open weights, Chinese labs key
By
–
The US frontier labs have all walked away from open weights. They continue to occasionally release excellent open models (Gemma 4, etc), but they are smaller models that are not competitive with their closed weights models. So all eyes are on Chinese AI labs for open models.
-
Open questions about RSI, LLM gains, Chinese models, and open weights
By
–
Lots of open questions: is RSI indeed happening at the Big Three labs? How long can the exponential of LLM ability gain last? How much do Chinese models rely on distillation, and can they keep pace given chip constraints? Will there continue to be frontier open weights models?
-
Abacus AI Deep Agent: Full Stack Development with AI Automation
By
–
Vibe Code Full Stack And Mobile Apps – Use english to vibe code full stack software systems – vive code companion mobile apps – run large scale systems in production AI does everything for you Do it ALL on Abacus AI Deep Agent