---
title: Runloop
slug: runloop
url: "https://toolweight.com/tools/runloop"
homepage: "https://runloop.ai"
categories: sandbox-providers
last_verified: 2026-11-07
dataset_version: 0738be72
license: CC-BY-4.0
---

# Runloop

> Devboxes, blueprints and eval harnesses aimed at teams training coding agents.

- Persistent devboxes plus blueprints, snapshots and public benchmark harnesses for agent evals
- Narrower than E2B: built for SWE-bench-style evaluation loops and training-data generation, not general untrusted code
- For that workload the surrounding tooling is worth more than the boot time

## Identity

|  |  |
| --- | --- |
| Name | Runloop |
| Company | Runloop |
| One-liner | Devboxes, blueprints and eval harnesses aimed at teams training coding agents. |
| Site | https://runloop.ai |
| Docs | https://docs.runloop.ai |
| Founded | 2023 |
| Funding | Seed |
| Compared in | 1 |

## Where it is compared

### [Sandbox providers](https://toolweight.com/compare/sandbox-providers)
Ranked **#4 of 19** on default weights.
| Field | Value | Confidence | Verified | Source | Note | Permalink |
| --- | --- | --- | --- | --- | --- | --- |
| Cold start | 800ms | Inferred | - | - | Estimate for a devbox created from an existing blueprint or snapshot. | https://toolweight.com/compare/sandbox-providers#runloop--cold_start_measured |
| Cold start (claimed) | 1s | Vendor-claimed | 2026-01-15 | https://docs.runloop.ai | Marketed as sub-second devbox boot from a snapshot. Recorded at 1000 ms, the literal ceiling of "sub-second", which is this page's convention for an unquantified claim of that shape, the same reading Modal's identical phrasing gets. It was previously recorded as 500, a favourable mid-reading no other vendor received. | https://toolweight.com/compare/sandbox-providers#runloop--cold_start_claimed |
| Isolation | microvm | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--isolation_tech |
| Root in guest | ● | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--root_access |
| Egress control | open | Inferred | - | - | No documented per-devbox allowlist. | https://toolweight.com/compare/sandbox-providers#runloop--egress_control |
| Max runtime | 24h | Inferred | - | - | Devboxes are long-lived with a configurable idle shutdown. | https://toolweight.com/compare/sandbox-providers#runloop--max_runtime |
| Persistent FS | ● | Inferred | - | - | Devboxes persist across suspend and resume. | https://toolweight.com/compare/sandbox-providers#runloop--persistent_fs |
| Snapshot & fork | ● | Vendor-claimed | 2026-01-15 | https://docs.runloop.ai | Snapshots and blueprints are the core primitive. | https://toolweight.com/compare/sandbox-providers#runloop--snapshot_fork |
| Custom images | ● | Vendor-claimed | 2026-01-15 | https://docs.runloop.ai | Blueprints build from a Dockerfile or setup commands. | https://toolweight.com/compare/sandbox-providers#runloop--custom_images |
| GPU | ○ | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--gpu |
| Runtimes | python, javascript-typescript, bash, any-oci-image | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--language_runtimes |
| Preview URLs | ● | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--preview_urls |
| Browser inside | ◐ | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--browser_in_sandbox |
| File up/download | ● | Vendor-claimed | 2026-01-15 | https://docs.runloop.ai | - | https://toolweight.com/compare/sandbox-providers#runloop--file_transfer_api |
| Sydney region | ○ | Inferred | - | - | US only. | https://toolweight.com/compare/sandbox-providers#runloop--sydney_region |
| Concurrent limit | 100 sandboxes | Inferred | - | - | Plan-dependent; approximate. | https://toolweight.com/compare/sandbox-providers#runloop--max_concurrency |
| MCP server | ◐ | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--mcp_server |
| SDKs | python, typescript, rest-api, cli | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--sdk_languages |
| Streaming output | ● | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--streaming_stdout |
| Price / hour | $0.32/hr | Vendor-claimed | 2026-11-07 | https://runloop.ai/pricing | 2 vCPU x $0.00003 x 3600 = $0.216 plus 4 GB x $0.000007 x 3600 = $0.101, total $0.317/hr. Excludes storage. | https://toolweight.com/compare/sandbox-providers#runloop--price_per_hour |
| Meter | per-second | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--billing_granularity |
| Idle cost | storage-only | Inferred | - | - | Suspended devboxes bill storage. | https://toolweight.com/compare/sandbox-providers#runloop--idle_billing |
| GitHub stars | - | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--github_stars |
| Funding | seed | Community-reported | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--funding_stage |
| Shipped | 2024 | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--launched |
| Positioning | Infrastructure and evaluation harnesses for teams building coding agents. | Inferred | - | - | - | https://toolweight.com/compare/sandbox-providers#runloop--positioning |

**Verdict.** Built around evaluation, not production traffic: the obvious pick if you measure an agent's pass rate nightly.

## Alternatives

- [Blaxel](https://toolweight.com/tools/blaxel), Agent-first cloud claiming ~25 ms microVM boots from snapshots.
- [boat](https://toolweight.com/tools/boat), Persistent Ubuntu VMs for agents with SSH, Docker and disk-level fork, billed per second.
- [Cloudflare Sandbox](https://toolweight.com/tools/cloudflare-sandbox), Container sandboxes driven from a Worker, addressed through Durable Objects.
- [CodeSandbox SDK](https://toolweight.com/tools/codesandbox-sdk), Firecracker VMs with memory snapshots, from the online IDE, now owned by Together AI.
- [Daytona](https://toolweight.com/tools/daytona), Sub-second container sandboxes for agent workloads, from a team that built a dev-env manager.
- [Self-hosted Firecracker](https://toolweight.com/tools/diy-firecracker), The baseline: Firecracker on your own metal, plus every hard part you now own.
- [E2B](https://toolweight.com/tools/e2b), Open-source Firecracker sandboxes with Python and TypeScript SDKs for AI agents.
- [exe.dev](https://toolweight.com/tools/exe-dev), Persistent VMs you SSH into, with root, apt and systemd, on a flat monthly plan.
- [Fly.io Machines](https://toolweight.com/tools/fly-machines), Raw Firecracker microVMs with a REST API, durable volumes and 35+ regions.
- [Freestyle](https://toolweight.com/tools/freestyle), Run untrusted JavaScript and full dev servers, with git hosting and domains built in.
- [GitHub Codespaces](https://toolweight.com/tools/github-codespaces), Devcontainer-backed cloud VMs built for humans, occasionally repurposed for agents.
- [Islo](https://toolweight.com/tools/islo), Per-agent isolated cloud sandboxes with enterprise policy controls, from Incredibuild.

## Licence and attribution

Data from Toolweight (https://toolweight.com), licensed CC-BY-4.0.

- Licence: [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)
- Canonical HTML: https://toolweight.com/tools/runloop
- Machine-readable: https://toolweight.com/tools/runloop.md · https://toolweight.com/api/v1 · https://toolweight.com/mcp
- Toolweight takes no affiliate revenue and sells no placements.