Runner H is an AI computer-use agent platform from H Company that automates tasks across browsers, desktops and enterprise systems without integrations or rebuilds.
About
Runner H is the agentic platform built by H Company, a Paris-founded AI research lab, to bring what the company calls operational intelligence to the enterprise. Its agents operate software the way a person does — clicking, typing, and navigating through interfaces — so businesses can automate workflows that span desktop applications, browsers, mobile interfaces and documents without rebuilding their existing systems or waiting on custom integrations.
The platform is powered by H Company's own Holo family of multimodal models, including Holo4, which is trained specifically for computer use in enterprise environments. Because H Company controls the full stack from model training to inference, it can manage cost, performance and roadmap end to end. Customers can adopt the platform three ways: a full enterprise deployment with forward-deployed engineers on site, an Agents API for builders, or an OpenAI-compatible Models API for researchers who want to run Holo4 in their own agent loops.
Key Features
Pricing & Fees
- Pricing: Agents API plans are Free ($0, up to 3 concurrent agents, ~60 sessions/month), Builder ($20/month, up to 10 concurrent agents, ~1,000 sessions/month), Startup ($200/month, up to 100 concurrent agents, ~10,000 sessions/month), and Enterprise (custom pricing); Models API is billed per token, and Enterprise pricing is by sales contact.
- Free tier: A free plan and free trials of HoloTab, HoloDesktop CLI and the open Hugging Face weights let users start at no cost.
- Compliance: The company states it is SOC 2 Type II, ISO 27001 and GDPR compliant.
- Data retention: The Models API is listed with zero data retention.
Who It's For
Runner H is aimed at enterprise teams that need to automate fragmented workflows across legacy desktop software, browsers and documents without re-engineering their stack. Typical users include operations managers automating claims or expense processing, IT and automation engineers building agents through the Agents API, and AI researchers experimenting with Holo4 through the OpenAI-compatible Models API or open weights. Startups and individual builders can start on the free or Builder tiers, while larger organizations get forward-deployed engineers and negotiated enterprise terms.
FAQ
What is Runner H?Runner H is the AI agent platform from H Company that uses computer-use agents powered by its Holo4 models to operate existing software — clicking, typing and navigating browsers and desktops — without integrations or system rebuilds.
Does Runner H have a free plan?Yes. The Agents API has a free tier with up to 3 concurrent agents and roughly 60 sessions per month, and HoloTab, HoloDesktop CLI and open model weights on Hugging Face are free to try.
How much does Runner H cost?Paid Agents API plans are Builder at $20/month and Startup at $200/month, with custom Enterprise pricing by sales; the Models API is charged per token and top-ups on paid plans carry a 20% markup over model cost.
Does Runner H require APIs or integrations to automate my software?No. The agents work directly in a cloud-hosted browser or on a desktop by interacting with the user interface itself, so existing applications can be automated without APIs or rebuilds.
How does H Company handle data privacy?The company states it is SOC 2 Type II, ISO 27001 and GDPR compliant, and the Models API is offered with zero data retention.
Related Tutorials
View all →
Related Tools
View all →
Open Agents - Cloud Coding Agents Spawn coding agents that run infinitely in the cloud. Powered by AI SDK, Gateway, Sandbox, and Workflow SDK. 
Emergence AI Emergence AI builds governed, verified agentic infrastructure that lets enterprises run autonomous AI agents safely across mission-critical systems like semiconductor manufacturing. 
Skydive Skydive is a cloud platform for building AI agents that run on their own computer with a real browser and carry out tasks across your team's existing tools. 
Yuxi - RAG Harness for AI Agents Yuxi provides a RAG harness to enhance AI agents with retrieval-augmented generation capabilities. 
OmniRoute - Free AI Gateway for Multi-Provider LLMs Free, open-source AI router with auto-fallback. 236 providers, one endpoint, 95 MCP tools, 17 routing strategies, A2A protocol, auto-combo engine, semantic cache, memory & skills. Deploy anywhere.