Gradient
2 employees on LinkedIn · 6 known investors
Gradient AI provides a fleet of agents that research, build, and deliver digital content and applications—including apps, websites, documents, and video—using the customer's data, tools, and style.
Also known as Hyperagent
Founders & leadership
Investors · 6
Company profile
researched Aug 2026Gradient, operating under the product name Hyperagent, offers a service in which a user assigns a brief and a fleet of AI agents researches, drafts, builds and delivers the resulting work. The company describes outputs as "deliverables that stay alive" — artifacts that the agents continue to update as the underlying information changes, rather than one-off generations. Supported output formats listed on the company site include fully functional websites (HTML built and hosted rather than pasted into a document), video, slide decks, documents, dashboards and applications, alongside examples such as market signal briefs, candidate slates, comp analyses and landing pages.
Each agent is provisioned with its own computing environment, including browsers and shells, and can be connected to the customer's existing stack in order to search, write code, make decisions and produce artifacts. The site emphasizes transparent execution, stating that users can see every search, decision and artifact as it is built, and that Hyperagent generates new memories and skills as it works. Agents are deployed, connected and managed from a single command center.
Published example workloads on the site each carry a stated runtime and cost: an out-of-home campaign (billboard location research, creative strategy and real-world visualization) at 8 minutes and $6.41; a cinematic real estate video marketing package at 29 minutes and $24.79; a hiring command center that sources and ranks candidates in a dashboard with skill sliders at 22 minutes and $14.18; and a complete brand transformation covering identity, packaging and a launch site at 15 minutes and $3.88.
Business model
Gradient sells access to autonomous agents that produce and maintain work artifacts on behalf of customer teams. The company's site displays a per-task dollar figure alongside each example deliverable (ranging from $3.88 to $24.79), suggesting usage- or task-based cost accounting, but no pricing plans, contract terms or billing details are stated in the available sources.
▸Full profile — market position, technology, go-to-market, risks & controversies
Market position
The company positions the product against static AI output on three points stated on its site: deliverables that are kept current as the customer's context changes rather than produced once; agents that operate with real infrastructure (browsers, shells, access to the customer's stack) rather than only generating text; and transparent, inspectable execution in which each search, decision and artifact is visible. It also emphasizes producing directly usable formats, such as live hosted HTML websites rather than code pasted into a document.
Technology
The product is built around agents that each run in their own computing environment with access to real browsers and shells, enabling them to search the web, write and execute code, and connect to the customer's existing tools and data. The company states that agents record their execution steps for user inspection and accumulate memories and skills over time. Delivery integrations described include Slack @-mentions, Telegram, email, webhooks and scheduled runs.
Go-to-market
Distribution is oriented around meeting users inside tools they already use: agents can be invoked by @-mention in Slack, connected to Telegram, triggered via webhook or run on a schedule, with output delivered to Slack, email or elsewhere. The website offers a direct self-serve entry point where a visitor describes what they want Hyperagent to build.
Team-based business users who need recurring produced work across marketing, recruiting and analysis functions. Example use cases published by the company point to marketing and advertising (out-of-home campaigns, brand identity and packaging), real estate marketing, recruiting and hiring, and market or compensation research.
Risks & controversies
Several unrelated entities share the Gradient name — including the San Francisco AI seed venture firm at gradient.com, the London-based agentic AI company Gradient Labs, and a Japan/APAC go-to-market advisory called Gradient Consulting — which creates a material risk of identity confusion in databases and public coverage.
Compiled by commissioned research from 8 cited public sources — announcements, filings, and press listed under research sources below.
Key figures
latest reportedCompany-reported or press-reported figures, each dated to when it was claimed — not independently audited.
In the news
▸Research sources · 8
primary sources listed
- Gradientgradient.ai · web
8 public sources were cited for this profile; the first-party ones are listed here.
Frequently asked questions
- What does Gradient do?
- Gradient (gradient.ai) builds Hyperagent, a fleet of AI agents that research, build and maintain deliverables using a team's data and tools.
- Who founded Gradient?
- Gradient was founded by Eric Yang.
- Who are Gradient's investors?
- Gradient's investors include Ansa Capital, Wing, CommerzVentures, Exceptional Capital, Octopus Ventures, Redpoint Ventures.