Open source alternatives to Cognition Labs
200 open source apps like Cognition Labs, closest match first.
Top picks
- D
Devika
LaunchDevika is an open-source agentic AI software engineer. You give it a high-level instruction, and it breaks the task into steps, researches the relevant information on the web, and writes code to meet the objective. It is modeled on Devin, t
FreeOpen source
- U
Useagent
LaunchThe open-source Devin alternative. Run Claude Code, Codex or OpenCode in isolated cloud workspaces you can self-host.
FreeOpen sourceSelf-hosted
- G
- M
Maggy
LaunchWhat started as an opinionated Claude Code setup kit is now an autonomous AI engineering command center
FreeOpen source
- D
DeepSeek-Reasonix
LaunchOpen-source coding agent for your terminal, desktop, browser and editor. The program runs on your computer and uses the model account you configure. Describe a task; it reads your project, edits files and runs commands. You choose how often
FreeOpen source
- A
Autoresearch
LaunchAutoresearch is a skill for Claude Code, OpenCode and OpenAI Codex that runs an autonomous improvement loop toward a goal you set. It is based on the idea from Andrej Karpathy's autoresearch project: pick one measurable metric, keep the sco
FreeOpen source
More like Cognition Labs
Context Engineering Kit
Hand-crafted Claude Code Skills focused on improving agent results quality
About Cognition Labs
Cognition Labs
Cognition operates Devin, the first autonomous software engineer
Cognition makes Devin, an AI agent that works as an autonomous software engineer. Devin plans a task, writes the code, tests it and ships production code on its own, working inside your codebase and the tools your team already uses. The company describes its aim as helping engineers work more like architects, focusing on system design and problem solving while agents handle the repetitive engineering work. The site says Devin is deployed at some of the largest and most complex institutions in the world. Cognition also publishes FrontierCode, a benchmark that asks whether a human maintainer would merge a model's pull request. It scores code on correctness, test quality, scope discipline, style and adherence to codebase standards, not only on unit tests. The homepage links to a demo request and a Try Devin page, and the site has research and blog sections. Cognition calls itself an applied AI lab that builds agents and tools to expand human capacity.
- Category
- AI
- Website
- cognition.com

