Sidequest: we gave Holo4 a pile of bricks and a pile of blocks 🧱 Type an idea and watch it build. HoloBricks makes brick models. HoloBlocks makes worlds you can walk around in. Both built on our Agents API and open source. Code and demos below ⬇️

Today we're releasing Holo4, our new family of generalist computer-use models. Holo4 can click, code, and call tools across desktop, web, Android, code sandboxes, MCP servers, and business APIs. Available today on the H Models API for commercial use.

Agents are benefiting from frontier model intelligence, but are now blocked by search performance. So we developed NeoMME-Retriever, an efficient visual document retriever for Visual RAG. We contribute two models: • NeoMME, a family of 260M and 800M multimodal-native, multilingual foundation encoders trained from scratch. One bidirectional Transformer processes raw image patches and text tokens. • NeoMME-Retriever, a visual document retriever built on the NeoMME backbone. It encodes PDF page as images without OCR and returns dense and late-interaction embeddings in one pass. Both retriever model sizes sit on ViDoRe v3's model-size Pareto frontier. With matched 2048x2048 inputs on one NVIDIA L40S, the 260M model encodes about 51 pages per second, nearly 2x ColModernVBERT's 26 pages per second. Built at @hcompany_ai by @tonywu_71 and @Aurelien_L_. We released the weights under Apache 2.0 with day-0 Hugging Face Transformers support. Read @tonywu_71's thread for more details. 👇

Got plans on the 17th? @hcompany_ai's throwing the afterparty for @dotaiconf No slides, only drinks, food and a DJ 🕺 With @dotConferences and Eugenia School 📅 Sep 17 🎟️ Grab your spot before they're gone: