A 12-month time difference between Gemma 3 27b and Gemma 4 31b. The jump is absolutely enormous. Just look at the evaluations between the two models. GPQA doubled, AIME 2026 went from ~20% to ~90%, and so on. Crazy.
@kimmonismus
-
Gamma 31b Unexpectedly Outperforms Qwen 3.5 397B Model
By
–
Gamma 31b model outperforming Qwen 3.5 397B is nuts to me.
-
Google Releases Gemma 4 Open-Source Models with Impressive Performance
By
–
Here we go: Gamma 4 released: ""Outperforms models 20x its size" Google dropped Gemma 4 under Apache 2.0, full open-source, big licensing shift. Built on Gemini 3 tech, four sizes: E2B, E4B, 26B MoE, 31B Dense. Price-performance: 31B is #3 open model on Arena AI, 26B MoE is #6
-
Genspark Offers Unlimited AI Chat and Image Access in 2026
By
–
5/ The wildest part? Genspark is offering unlimited usage of AI Chat and AI Image for all of 2026. Nano Banana 2, GPT Image, Flux, Seedream, Gemini 3.1 pro, GPT-5.4, Claude Opus 4.6 and more are available with unlimited access for paid users. In the web version click the blue
-
Testing Genspark Claw for Media Company Planning and Telegram Integration
By
–
4/ Here’s what I tried: I dropped in one prompt with everything I needed for a media company idea and let Genspark Claw work through it
— Chubby♨️ (@kimmonismus) 2 avril 2026
It mapped things out step by step and turned it into something I could actually follow
I also tested it inside Telegram and it works directly… pic.twitter.com/hbRtniCg6j4/ Here’s what I tried: I dropped in one prompt with everything I needed for a media company idea and let Genspark Claw work through it It mapped things out step by step and turned it into something I could actually follow I also tested it inside Telegram and it works directly
-
Using AI to generate and edit full presentations
By
–
3/ It turned everything into a full presentation… structure, visuals, and branding already there. I was able to open and edit it right away which I honestly didn’t expect, this makes my everyday stuff way easier to work through
-
Combining Multiple LLMs for Enhanced Workflow Utility
By
–
2/ What I really really liked is that it’s not just one model… It’s combining models like GPT, Claude, and Gemini in the background so you actually get finished output instead of just text, and i think that makes it actually useful
-
User experience with Genspark Claw AI agent workflow
By
–
1/ Tried Genspark Claw from @genspark_ai and it’s actually better than I expected
— Chubby♨️ (@kimmonismus) 2 avril 2026
I used it to figure out a project and instead of just giving answers, it followed a workflow and walked me through everything step by step while helping me put everything together into something I… pic.twitter.com/tOD6cE8K0r1/ Tried Genspark Claw from @genspark_ai and it’s actually better than I expected I used it to figure out a project and instead of just giving answers, it followed a workflow and walked me through everything step by step while helping me put everything together into something I
-
Neuralink BCI technology restores speech for ALS patients
By
–
Neuralink has given someone with ALS back their (own) speech. Brain implants will be the next big thing after robotics, and I'm ready for it!pic.twitter.com/RfLTYWmuLL
— Chubby♨️ (@kimmonismus) 2 avril 2026Neuralink has given someone with ALS back their (own) speech. Brain implants will be the next big thing after robotics, and I'm ready for it!
-
PrismML’s 1-bit Bonsai Models Achieve 10x Intelligence Density Improvement
By
–
Pretty insane: PrismML has introduced its 1-bit Bonsai models, compressing an 8.2B-parameter LLM into just 1.15 GB while maintaining performance comparable to leading 8B models. By focusing on “intelligence density,” the company claims roughly a 10x improvement in capability per