Vercel Brings Eve agents Into Its Deployment Dashboard.

Vercel is moving AI agent development into the same dashboard developers already use to deploy web applications. The company now lets users create and launch Eve agents without first assembling a repository, configuring a project, or wiring up a separate chat interface.

According to Vercel’s changelog announcement, the new builder produces a working agent that users can chat with immediately. Behind the streamlined setup, it scaffolds the agent’s files, creates a private Git repository, and deploys the result as a new project on the user’s Vercel team.

That combination is the important part. The dashboard is not simply generating a disposable prototype. It creates a conventional, code-backed project that developers can inspect and customize after the initial deployment.

The builder starts with instructions that define the agent’s identity and behavior. Users can then choose any model available through Vercel’s AI Gateway, rather than being restricted to a single provider or model family.

Vercel says users can configure several core elements from the dashboard:

• Agent instructions: Define the agent’s role, responsibilities, and expected behavior.
• Model selection: Pick from models available through AI Gateway.
• User interface: Add a Next.js web chat or a Slack channel where people can interact with the agent.
• Tools and data: Connect services including Linear and Notion, or provide access through a custom MCP server.
• Deployment and source code: Generate a private Git repository and deploy it as a Vercel project on the selected team.

Users can begin the process by selecting Add New → Agent in the Vercel dashboard. The result is a deployed endpoint and interaction layer, with the underlying repository available for further development.

Agent projects often require several pieces of infrastructure before a team can test the actual idea. Developers may need to select a model, create API routes, build a chat interface, configure credentials, connect external data, establish a repository, and set up deployment. Vercel’s builder compresses those steps into a guided workflow while retaining access to the generated code.

For small businesses and product teams, that could make Eve agents useful for quickly testing internal assistants, customer-facing support experiences, knowledge tools, or agents that take actions in connected systems. A team could begin with a web chat, for example, then modify the generated Next.js application as requirements become clearer. Slack support also provides a route to internal adoption without asking employees to learn another interface.

The connections are particularly relevant because useful agents generally need more than model access. They need permission to retrieve information or perform actions in business systems. Vercel’s built-in integrations and custom MCP server option provide two paths: a simpler setup for supported services, and a more flexible route for proprietary tools or less common platforms.

The launch also fits Vercel’s broader effort to assemble an agent development stack around its deployment platform. The dashboard builder relies on AI Gateway for model access, while Vercel Connect is intended to provide scoped access to external services. Connect recently reached general availability with scoped agent access. Vercel has also added per-user AI Gateway budgets, giving teams another mechanism for controlling model expenditure.

There are still details teams will need to evaluate independently. Vercel’s brief announcement does not specify separate pricing for the Eve agent builder, usage limits, supported Git providers, or a complete list of built-in connections. It also does not remove the need to review an agent’s permissions, instructions, and generated code before using it with sensitive business data.

Still, the update reduces the distance between an agent idea and a deployed application. For existing Vercel users, the new workflow turns agent creation into another project type rather than a separate infrastructure exercise.

What is Vercel Eve agents? Eve agents are AI-powered agents that users can create and deploy from the Vercel dashboard. The builder generates the agent’s files, creates a private Git repository, deploys a Vercel project, and provides an interface for chatting with the resulting agent.

What interfaces and integrations do Eve agents support? Vercel says users can add a Next.js web chat or a Slack channel. Agents can receive tools and data through built-in connections such as Linear and Notion, or through a custom MCP server.

How can users create an Eve agent in Vercel? Users can select Add New → Agent from the Vercel dashboard and follow the builder. Vercel says the capability is available now, but its announcement does not provide separate pricing or plan requirements.