# Devin review

Devin is Cognition's autonomous software engineer. It writes and tests in an isolated Devbox and opens a PR on your SCM. The Brain always stays in Cognition's cloud. This page is a review with dates, not a score out of ten.

Facts checked 15 September 2026. Prices and limits move. Read the citations.

Rush competes for the same job. Facts below are from Cognition docs, Cognition blogs, and named press. Independent numbers are labeled.

## The useful distinction

Devin is Cognition's autonomous software engineer. It writes and tests in an isolated Devbox and opens a PR on your SCM. The Brain always stays in Cognition's cloud.

This page is a review with dates, not a score out of ten. Cognition's 67% merge rate is from 14 November 2025. Answer.AI's 3-out-of-20 is from 8 January 2025.

## Where Rush sits

If you want Devin's job with Claude Code, Codex, or OpenCode as the engine, that is Rush. If you want Cognition's agent, stay on this page.

## The field

| Tool | Job | What runs | Isolation | Proof | Price | Taps out |
| --- | --- | --- | --- | --- | --- | --- |
| Rush | Send a task by API; an agent does it in its own sandbox | Claude Code, Codex or OpenCode | One E2B microVM per run | PR with checks, artifacts, transcript; else needs_review | Free: a one-time $1 free credit, then Pro $40 per member | Three harnesses. GitHub repos only. No self-host. |
| Devin | Cognition's autonomous engineer on a ticket | Devin Brain + Devbox | Isolated VM per session | PR on your SCM | Pro $20 / mo; Teams $80 min | No BYO model keys. Opaque quotas. |
| Cursor Cloud Agents | Cursor Agent, unattended, in a cloud VM | Cursor Agent | Firecracker microVM | Draft PR + artifacts | Paid Cursor plan, from $20 / mo | Needs Cursor. Commands auto-run. |
| GitHub Copilot cloud agent | Issue-to-PR inside GitHub Actions | Copilot cloud agent | Ephemeral Actions VM | Draft PR + session logs | Paid Copilot, from $10 / mo | 59-minute cap. GitHub repos only. |

## The file

### What Cognition says it is for

Vendor review 14 Nov 2025.. Tickets you can do in about three hours, with verifiable success criteria.

Official when-to-use: if you can do it in three hours, Devin can most likely do it. Break longer work into sessions.

The 14 November 2025 vendor review: senior-level at codebase understanding, junior at execution. Vendor metrics from that post: 67% of PRs merged versus 34% the year before. Treat those as Cognition's numbers.

**Taps out.** Ambiguous senior-scope work. Mid-task requirement changes. Unverifiable outcomes without a human.

### Pricing, as of 15 Sep 2026

devin.ai/pricing and docs.self-serve. 15 September 2026.. Individuals on Pro $20 or Max $200. Teams at $80 minimum.

14 April 2026: Cognition retired Core and the $500 Team plan. Self-serve extra usage moved from ACUs to dollars. Enterprise stayed on ACUs.

Billing docs and the marketing page disagree on Teams member caps (200 vs unlimited) and on concurrent session limits. Record both. Do not reconcile them here.

**Taps out.** Exact Pro/Max quota sizes are unpublished.

### Independent evaluation

Answer.AI 8 Jan 2025.. A dated baseline, not a 2026 score.

Answer.AI (8 January 2025): 20 real tasks, 14 failures, 3 successes, 3 inconclusive. Quote: Devin would spend days pursuing impossible solutions rather than recognizing fundamental blockers.

The Register (23 January 2025) and TechCrunch (3 April 2025) repeated that result. Devin 2.0 shipped 3 April 2025. Do not treat 3/20 as current.

**Taps out.** No fetched 2026 independent re-run of that protocol.

## The off-ramp

1. Code that Devin pushed to GitHub, GitLab, Bitbucket, or Azure DevOps stays on that remote.
2. Uninstall the Devin GitHub App from github.com/settings/installations.
3. Session messages and attachments can be pulled from the v3 API. There is no documented one-click export of every cloud session.
4. Privacy requests: privacy@cognition.ai. Paid plans can opt out of training. Enterprise trains only with written consent.

## Skip Rush if

- You need BYO model keys. Official docs: not supported.
- You need a current independent win-rate. The best third-party study is January 2025.

## Questions people actually ask

### Is Devin SOC 2?

Cognition's security docs say SOC 2 Type II was obtained in March 2024, plus ISO 27001. Rush does not make that claim for itself.

### Can I self-host the Brain?

No. Outposts move execution. The Brain stays in Cognition's cloud.

## Keep reading

- [Devin alternatives](https://getrush.ai/compare/devin-alternatives.md)
- [Rush vs Devin](https://getrush.ai/compare/rush-vs-devin.md)
- [How to move off Devin](https://getrush.ai/compare/move-from-devin.md)
- [Devin vs Cursor Cloud Agents](https://getrush.ai/compare/devin-vs-cursor.md)

## Run the next task on Rush

### Rush

- [How Rush runs each agent in its own cloud computer](https://getrush.ai/cloud/)
- [Rush plans and pricing](https://getrush.ai/pricing.md)
- [Quickstart: start a run with an API key](https://getrush.ai/docs/quickstart.md)
- [Review every pull request](https://getrush.ai/docs/use-cases/review-pull-requests.md)

### Connect your tools

- [Run agents with Linear](https://getrush.ai/integrations/linear.md)
- [Run agents with Sentry](https://getrush.ai/integrations/sentry.md)
- [Run agents with GitLab](https://getrush.ai/integrations/gitlab.md)
- [Run agents with Slack](https://getrush.ai/integrations/slack.md)
- [40 more apps and MCP servers Rush connects to](https://getrush.ai/integrations.md)
- [Hand work to Rush from any MCP client](https://getrush.ai/docs/concepts/mcp-server.md)

### Run it on Rush

- [Review Pull Requests Automatically in a Next.js App](https://getrush.ai/run/review-pull-requests/nextjs.md)
- [Review Pull Requests on a Python Library](https://getrush.ai/run/review-pull-requests/python.md)
- [Review Pull Requests for Accessibility in a React App](https://getrush.ai/run/review-accessibility/react.md)
- [Review Pull Requests Automatically in a Go Service](https://getrush.ai/run/review-pull-requests/go.md)

## Sources

1. Cognition Docs, "Introducing Devin", retrieved 2026-09-15. https://docs.devin.ai/get-started/devin-intro
2. Cognition, "Plans and Pricing | Devin", retrieved 2026-09-15. https://devin.ai/pricing
3. Cognition Docs, "When to Use Devin", retrieved 2026-09-15. https://docs.devin.ai/essential-guidelines/when-to-use-devin
4. Cognition, "Devin's 2025 Performance Review", published 2025-11-14. https://cognition.com/blog/devin-annual-performance-review-2025
5. Cognition Docs, "Enterprise Deployment", retrieved 2026-09-15. https://docs.devin.ai/enterprise/deployment/overview
6. Cognition Docs, "Security at Cognition", retrieved 2026-09-15. https://docs.devin.ai/admin/security
7. Answer.AI, "Thoughts On A Month With Devin", published 2025-01-08. https://www.answer.ai/posts/2025-01-08-devin
