Hamza Benchekroun

@m_sirovatka @rasdani_ Congrats! Awesome work

They distilled me into a blogpost today 🫡 was a lot of fun optimizing the whole stack over the last few months, still a long way to go but the whole team cooked on this release 🚀

@hcompany_ai @nvidia 🚀

H Company joins the @nvidia Nemotron Coalition! The coalition brings together leading AI labs working to advance open frontier models. As frontier models continue to evolve, the conversation is expanding beyond reasoning and generation to a new challenge: Action. The ability for AI systems to understand interfaces, navigate software, and complete real-world tasks is becoming a key part of the next frontier. At H, this is the challenge we tackle every day through computer-use AI. We're proud to contribute this perspective alongside the coalition's members and help advance open models. And proud to see a European company at the table. Read more in the comments.

HoloTab is officially live. 🚀 14 days ago, we released Holo3. It proved you can hit state-of-the-art computer-use performance, outperforming the biggest models in the world at 1/10th of the cost. Now, we are moving that power out of the research labs and into your browser tabs. Don't trust the benchmarks. Trust your own experience! #AI #AgenticAI #HoloTab #ComputerUse #Innovation #HCompany

When Jensen Huang tells you that computer-use is extremely hard, and that what you are building stands out, it means something. We had the chance to sit down with him and the @nvidia team for a candid, in-depth conversation, coming right after the release of #Holo3. We showed them something we are almost ready to share. Stay tuned! 📢 Thank you @Accel and @nvidia for organizing!

Holo3 is here 🚀. Today, we're launching Holo3: our new series of frontier computer-use models. 78.9% on OSWorld-Verified. That puts us ahead of GPT-5.4 and Opus 4.6, at one-tenth of the cost. Weights on Hugging Face. API is live. Test it now! #Holo3 #OpenSource #ComputerUse #OSWorld #AI #AgenticAI

@Alibaba_Qwen Great job! Congrats

🚀 Introducing the Qwen 3.5 Medium Model Series Qwen3.5-Flash · Qwen3.5-35B-A3B · Qwen3.5-122B-A10B · Qwen3.5-27B ✨ More intelligence, less compute. • Qwen3.5-35B-A3B now surpasses Qwen3-235B-A22B-2507 and Qwen3-VL-235B-A22B — a reminder that better architecture, data quality, and RL can move intelligence forward, not just bigger parameter counts. • Qwen3.5-122B-A10B and 27B continue narrowing the gap between medium-sized and frontier models — especially in more complex agent scenarios. • Qwen3.5-Flash is the hosted production version aligned with 35B-A3B, featuring: – 1M context length by default – Official built-in tools 🔗 Hugging Face: huggingface.co/collections/Qw… 🔗 ModelScope: modelscope.cn/collections/Qw… 🔗 Qwen3.5-Flash API: modelstudio.console.alibabacloud.com/ap-southeast-1… Try in Qwen Chat 👇 Flash: chat.qwen.ai/?models=qwen3.… 27B: chat.qwen.ai/?models=qwen3.… 35B-A3B: chat.qwen.ai/?models=qwen3.… 122B-A10B: chat.qwen.ai/?models=qwen3.… Would love to hear what you build with it.

Holo2-235B-A22B: #1 on ScreenSpot-Pro, #1 on OSWorldG 🎯 🚀 Today, we are releasing Holo2-235B-A22B 🤗, our most capable GUI localization model yet! Holo2 now leads on all major GUI localization benchmarks: 78.5% on ScreenSpot-Pro and 79.0% on OSWorld-G!

Our Surfer 2 tech report is out 🌊 Surfer 2 achieves SOTA on all major computer-use benchmarks and surpasses humans with retries. Report: assets.hcompanyprod.fr/surfer2_tech_r… Join us: jobs.ashbyhq.com/hcompany

@GFaang97609 no i never run ablations, because i believe god has a plan

🚀 Holo1.5 is here. Our next-gen open foundation models for Computer Use agents: 3B, 7B, and new 72B. ✨ +10% accuracy vs Holo1 💻 SOTA UI localization & understanding 🤗 Open weights on HuggingFace: huggingface.co/Hcompany/Holo1… 📝Blog post: hcompany.ai/blog/holo-1-5 More to come soon...

H company just released Holo1.5, new sota vision language action models for web navigation and localization 💻 it comes in 3B, 7B and 72B, and 7B comes in Apache 2.0 license 🔥

🚀 As the autonomy of AI agents keeps improving quickly, our team at @hcompany_ai is excited to share our latest model family: Holo1.5—VLMs that are SOTA on localization and UI QA tasks, two essential skills for reliable computer/mobile use! (1/4 🧵)

We’ve been cooking this summer: Holo1.5 is here! SOTA UI localization + QA, 3× gains vs Qwen-2.5 VL 🍳 Now up to 72B 💥 — a strong base for computer-use agents like Surfer. • Open weights on HuggingFace 🤗 huggingface.co/Hcompany/Holo1… • Blog post 📝 hcompany.ai/blog/holo-1-5 (1/n 🧵)

@andimarafioti 🫡 ❤️

Fuck it. Today, we open source FineVision: the finest curation of datasets for VLMs, over 200 sources! > 20% improvement across 10 benchmarks > 17M unique images > 10B answer tokens > New capabilities: GUI navigation, pointing, counting FineVision 10x’s open-source VLMs.

Thank you for the book @huggingface, time for GPUs to go brrrr 😍 (been cooking hard with @hparams at @hcompany_ai, stay tuned!)

@huybery Qwen3-VL when ? 👀

We've updated the thinking model, let's keep moving forward! There are more surprises on the way! 🚀

@cis_female 😂😂😂

> fp8 is 100 tflops faster when the kernel name has "cutlass" in it kms github.com/triton-lang/tr…

2️⃣ Holo-1: We are Open-Sourcing our Visual-Language Model powering Surfer H We’ve beefed up Surfer H’s web automation with Holo-1, our 3B & 7B-parameters Action Models. It now achieves industry-leading UI localization and navigation accuracy while staying compact and cost-efficient. We’re releasing Holo-1 plus the 1,639-scenario WebClick dataset on Hugging Face to spark the next wave of visual-AI innovation. Explore Holo-1 & WebClick now on : Hugging Face : huggingface.co/Hcompany Amazon Marketplace: aws.amazon.com/marketplace/se…

Today, we’re thrilled to announce 3 major steps forward in bringing our vision of Agentic AI to life: 1️⃣ Runner H : Public Beta is now live! Imagine having your AI agent execute entire workflows across web apps, documents, spreadsheets, and more with a single prompt.

Based 🫡