Image created with gemini-2.5-flash-image with claude-sonnet-4-5-20250929. Image prompt: A cinematic photograph of an elegant two-layer birthday cake on a residential doorstep at golden hour, house numbers visible, warm neighborhood street with glowing porch lights in soft focus behind, rich reds and deep blues in the frosting against slate grey concrete, a small card reading ‘104’ beside the cake, high contrast lighting, modern and stylish composition.
Two years ago. No model had surpassed GPT-4 & it wasn’t clear that was possible. Now you can get better than GPT-4 level performance on open weights models running on consumer hardware, and the state of the art in LLMs is cheaper & faster and very much more capable than GPT-4.”” / X https://x.com/emollick/status/1970790843868213361
EmbeddingGemma: lightweight SOTA embeddings • 308M params, built on Gemma 3 • Tops MTEB (multilingual, English, code) <500M models • Matches models 2× its size, efficient even w/ 4-bit or 128-dim embeddings • Encoder-decoder init + geometric distillation • Spread-out https://x.com/arankomatsuzaki/status/1971041110446465251
Introducing Magistral Small 1.2 & Magistral Medium 1.2, minor updates to our Magistral 1.1 models! – Multimodality: Now equipped with a vision encoder, these models handle both text and images seamlessly. – Performance Boost: 15% improvements on math and coding benchmarks such https://x.com/MistralAI/status/1968670593412190381
Beautiful open smol moe for vision language task Weight are here if you want to try: https://x.com/eliebakouch/status/1968809452640825650
Modular’s GenAI stack now provides the world’s best Blackwell and MI355X performance, as well as developer support for Apple/NV/AMD consumer GPUs. Never before has it been this easy to get going, and unified pip package gives you choice of how to deploy! Check it out👇🚀”” / X https://x.com/clattner_llvm/status/1970203319097495891
