Releases·Model Release
H Company releases Holo4-27B and Holo4-35B-A3B vision-language models for Computer Use
H Company released Holo4, vision-language models for Computer Use, on Hugging Face in GGUF format. Two variants ship: a 27B dense model built on Qwen3.8 and a 35B-A3B mixture-of-experts model built on Qwen3.5. Paired with the hai-agents harness, they take screenshots and tool results as input and execute clicks, typing, code, and tool calls.
26 engagement2 sourcesMon, Sep 28, 2026, 10:59 AM
AI Overview
- H Company released Holo4-27B-GGUF, a Computer Use VLM built on Qwen3.8 dense architecture, on Hugging Face
- Holo4-35B-A3B-GGUF is a 35B Computer Use VLM built on the Qwen3.5 mixture-of-experts architecture
- With the hai-agents harness, the models receive screenshots and tool results, then execute clicks, typing, code, and tool calls
Related
Social·Community
H Company releases Holo4, an open-weight 27B agent model that clicks GUIs, runs code, and calls MCP tools
A YouTube video by Sam Witteveen reviews Holo4, H Company's open-weight model family that operates a GUI, writes and runs code, and calls MCP tools within one agent loop. The video links H Company's blog and Hugging Face weights, showcasing the release to developers.
A YouTube video by Sam Witteveen reviews Holo4, H Company's open-weight model family that operates a GUI, writes and runs code, and calls MCP tools within one agent loop. The video links H Company's blog and Hugging Face weights, showcasing the release to developers.
384 engagement1 sourceyoutubeTue, Oct 6, 2026, 02:10 PM
Social·Community
How to run Holo4 on your own computer in six steps, and what its 85.2% score hides
Guide walks through running Holo4 locally and questions its 85.2% headline score
Guide walks through running Holo4 locally and questions its 85.2% headline score
0 engagement1 sourcerssMon, Oct 5, 2026, 01:27 PM
Releases·Model Release
Tencent releases UI-Mate-27B, an open-weight GUI agent for long-horizon desktop tasks
Tencent released UI-Mate-27B, an open-weight foundation GUI agent that operates across applications and operating systems by observing live screenshots and producing structured keyboard and mouse actions. It supports general computer use from natural-language instructions and demonstration-guided computer use, where a reusable workflow is extracted from one successful demonstration. The model is available on Hugging Face.
Tencent released UI-Mate-27B, an open-weight foundation GUI agent that operates across applications and operating systems by observing live screenshots and producing structured keyboard and mouse actions. It supports general computer use from natural-language instructions and demonstration-guided computer use, where a reusable workflow is extracted from one successful demonstration. The model is available on Hugging Face.
1 engagement1 sourceredditTue, Aug 18, 2026, 06:42 AM
Social·Community
Image Decision Models for RPA: Forms, Scans and Screenshots
A YouTube walkthrough by Sam Witteveen covers two open image decision models, ImaJev 4B and Jev-Omni, aimed at RPA decision tasks like checking screenshots, photos, and forms. The video explains why prior custom deep learning approaches fell short and demonstrates a Form Decision Inspector with confidence thresholds per question. Both models are available on Hugging Face, making them practical options for developers adding decision logic to automation pipelines.
A YouTube walkthrough by Sam Witteveen covers two open image decision models, ImaJev 4B and Jev-Omni, aimed at RPA decision tasks like checking screenshots, photos, and forms. The video explains why prior custom deep learning approaches fell short and demonstrates a Form Decision Inspector with confidence thresholds per question. Both models are available on Hugging Face, making them practical options for developers adding decision logic to automation pipelines.
469 engagement1 sourceyoutubeFri, Oct 2, 2026, 02:30 PM
Releases·Model Release
CohereLabs releases North Micro Vision Instruct, a 2.4B open-weight VLM
CohereLabs released North Micro Vision Instruct, a 2.4B-parameter open-weight vision-language model under Apache 2.0. It supports native-resolution images, multilingual and multi-image inputs, and is designed for prototyping and fine-tuning.
CohereLabs released North Micro Vision Instruct, a 2.4B-parameter open-weight vision-language model under Apache 2.0. It supports native-resolution images, multilingual and multi-image inputs, and is designed for prototyping and fine-tuning.
122 engagement1 sourceredditWed, Aug 12, 2026, 04:50 PM