---
type: "post"
id: "scale-ai-s-agentenv-agent-evals-can-now-f2ea3e"
slug: "scale-ai-s-agentenv-agent-evals-can-now-f2ea3e"
title: "Scale AI's AgentEnv agent evals can now run on Vercel Sandbox"
description: "Vercel announced that AgentEnv, Scale AI’s framework for testing agents against tasks and scoring their results, can now run on Vercel Sandbox."
retrieval_nugget: "Vercel announced that AgentEnv, Scale AI’s framework for testing agents against tasks and scoring their results, can now run on Vercel Sandbox."
published_at: "2026-10-26"
updated_at: "2026-10-26"
record_date: "2026-10-26"
date_kind: "published_at"
topics: ["evals"]
entities: ["Vercel","Vercel Sandbox","Scale AI","AgentEnv","Vercel CLI"]
related_patterns: ["agent-security-moves-to-runtime-boundaries","verification-bandwidth-is-the-scarce-resource"]
novelty_label: "new_fact"
what_changed: "Scale AI’s AgentEnv evaluations now run on Vercel Sandbox, distinct from the prior Harbor eval integration."
prior_refs: ["memory:02cd10d2-d648-8322-aa48-9fbd144c406a","memory:16ea1491-86aa-84a1-ac2b-25f177f4b358"]
source_url: "https://vercel.com/changelog/run-scale-ais-agentenv-evals-on-vercel-sandbox"
source_urls: ["https://vercel.com/changelog/run-scale-ais-agentenv-evals-on-vercel-sandbox"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:scale-ai-s-agentenv-agent-evals-can-now-f2ea3e"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/scale-ai-s-agentenv-agent-evals-can-now-f2ea3e/","markdown":"https://newruntime.com/posts/scale-ai-s-agentenv-agent-evals-can-now-f2ea3e.md","json":"https://newruntime.com/posts/scale-ai-s-agentenv-agent-evals-can-now-f2ea3e.json"}
---

# Scale AI's AgentEnv agent evals can now run on Vercel Sandbox

## Retrieval answer

Vercel announced that AgentEnv, Scale AI’s framework for testing agents against tasks and scoring their results, can now run on Vercel Sandbox.

Vercel announced that AgentEnv, Scale AI’s framework for testing agents against tasks and scoring their results, can now run on Vercel Sandbox. The Vercel provider ships with AgentEnv version 0.9.1334 and later.

The integration runs the agent and its task environment in separate Sandboxes. Vercel has published an example repository that records the task score and the agent’s actions. After completing setup, users can invoke an evaluation with `./vercel-eval --scope YOUR_TEAM --project YOUR_PROJECT`; the command uses an existing Vercel CLI login.

The announcement describes a provider integration and an example workflow, rather than a new AgentEnv evaluation method. It does not establish pricing, runtime performance, the range of supported tasks, service limits, or whether results differ from AgentEnv runs on other providers.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Evals](https://newruntime.com/topics/evals.md): Explore the evals topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)