Core
$20/mo
Plus $2.25/ACU
The autonomous AI software engineer from Cognition. Runs in its own cloud sandbox with an IDE, shell, and browser, takes a ticket from Linear, Jira, or Slack, plans the work, writes the code, runs tests, and opens a pull request — then iterates on review feedback. Devin 2.0 added Interactive Planning, Devin Search, Devin Wiki, and parallel Devins.
+ Pros
- Fully autonomous agent that handles the entire software development lifecycle end-to-end: from reading a Linear/Jira ticket, planning, navigating the codebase, writing code, running tests, and opening a pull request, then iterating on review feedback.
- Massive enterprise adoption and validated ROI: deployed at Citi, Mercedes-Benz, Goldman Sachs, Nubank, Dell, Santander, Infosys, Cognizant, and U.S. military branches. Nubank reported 8-12x engineering efficiency gains and 20x cost savings on ETL migrations.
- Multi-agent orchestration and parallelism: Devin can spin up a team of managed Devins that work in parallel with isolated VMs, coordinate tasks, resolve conflicts, and compile results — enabling multi-week, multi-repo projects.
- Extensive enterprise security and governance: SOC 2 Type II and ISO 27001 compliant, SAML/OIDC SSO, VPC deployment, zero-retention policy on customer code, RBAC, audit logs, secrets management, and network policy controls.
- Broadest integration ecosystem in the agent space: native integrations with GitHub, GitLab, Bitbucket, Linear, Jira, Slack, Microsoft Teams, Datadog, Sentry, Snowflake, AWS, Azure, GCP, plus an MCP marketplace with dozens of servers.
- Self-improving over time through fine-tuning and knowledge retention: Devin learns from past sessions, builds its own scripts and tools mid-migration, and uses playbooks and knowledge bases to get faster and more reliable on repeated tasks.
− Cons
- Usage-based pricing with ACUs (Agent Compute Units) can become expensive at scale. While Pro is $20/month and Teams $80/month, enterprise usage costs are custom and can escalate quickly for teams running many concurrent sessions.
- Not open source and fully dependent on Cognition's cloud infrastructure. There is no self-hosted option beyond VPC deployment for enterprise, and users cannot inspect, fork, or modify the agent's core code.
- Performance depends heavily on prompt quality and task scope. Devin itself warns that tasks under 3 hours work best — extremely complex or underspecified tasks can burn ACUs without delivering results.
Pricing
Team
$500/mo
250 ACUs included
Enterprise
Custom
VPC, SAML SSO, audit logs
Devin, created by Cognition, is the autonomous AI software engineer that defined the category. Unlike copilots and IDE assistants, Devin runs in its own cloud sandbox with an IDE, shell, and browser, and takes ownership of a task from start to finish: it reads a Linear or Jira ticket, plans the work, navigates the codebase, writes the code, runs the tests, and opens a pull request. After Devin 2.0 (April 2025), the platform shifted to usage-based pricing and added Interactive Planning, Devin Search, Devin Wiki, and parallel Devins in a shared cloud IDE.
In July 2025 Cognition acquired Windsurf and is converging both products into a single autonomous engineering platform — with Windsurf as the developer-facing IDE and Devin running in the background. Devin is deployed at large enterprises (Nubank, Dell, Cisco, and others), with usage-based pricing built on Agent Compute Units (ACUs).
Key Features
• Fully autonomous: takes a ticket and ships a PR without per-step guidance
• Cloud sandbox with IDE, shell, and browser per agent
• Interactive Planning: review and edit Devin's plan before execution
• Devin Search: natural-language queries across the codebase
• Devin Wiki: auto-generated, continuously updated architecture docs
• Parallel Devins: spin up multiple agents on different tasks at once
• Integrations: Linear, Jira, Slack, GitHub, GitLab, AWS, Azure, GCP, Snowflake
• Usage-based billing in ACUs (≈15 min of active work per ACU)
• Enterprise: VPC deployment, SAML SSO, audit logs, zero retention
Version History
Major update adding end-to-end desktop testing with computer use (Linux desktop access), Devin Review Autofix (self-review and fix code before PR), 3x faster startup time, and a redesigned UI unifying planning, coding, and review.
Introduced confidence ratings (🟢🟡🔴) that predict task success probability at multiple points per session, with automatic confidence scoring for Linear/Jira issues. Added built-in codebase intelligence (DeepWiki integration) for natural-language Q&A.
Major platform overhaul adding Interactive Planning (review/edit plans before execution), Devin Search (agentic codebase Q&A), Devin Wiki (auto-generated architecture documentation with diagrams), parallel Devins, and new pricing.
Devin can now run natively in Windows VMs, enabling building, testing, and iterating on .NET Framework applications, SQL Server workflows, Windows Forms, and other Windows-native development workflows with full computer use testing.
Launch of Devin CLI — run Devin agents directly from your terminal with full dev workflow support including planning, coding, testing, and PR creation.
Signature Snippet
Assign a Linear ticket "Migrate auth from Auth0 to Clerk across the monorepo" to Devin. Devin spins up a cloud workspace, reads the codebase, proposes a migration plan in Interactive Planning, runs the refactor across 100+ files, executes the test suite, fixes failures, and opens a draft PR — then iterates on your review comments.More in this Space
Sourcery
AI code review platform for the AI era. Automated code reviews, security scanning, and team analytics across GitHub, GitLab, VS Code, and JetBrains. Used by 300,000+ developers.
What The Diff
AI-powered PR description generator and code review assistant. Automatically writes pull request descriptions, sends stakeholder notifications, creates changelogs, and provides inline code refactoring.
Blackbox AI
Multi-agent AI coding platform with 12+ agents and 24+ models, featuring Chairman LLM for parallel multi-agent evaluation and end-to-end encrypted inference. Ships across six surfaces: CLI, IDE, Cloud, API, Mobile, and Builder.