Agents tagged coding

CLI-based multi-agent coding tool that uses a swarm of specialized agents in your terminal to plan, edit, and verify code in your repo.

Open-source code LLM and IDE assistant from Tsinghua/Zhipu AI offering autocomplete, chat, and translation across many languages and IDEs.

Autonomous AI software engineering agent for enterprise codebases, deployed via Slack/Linear/GitHub to handle real engineering tasks.

AI code review tool that automatically reviews pull requests on GitHub and GitLab with detailed line-by-line feedback. CodeRabbit is an AI-powered code review tool that automatically reviews pull requests on GitHub, GitLab, Azure DevOps, and Bitbucket.

Cloud development platform with AI features for instant dev environments and autonomous agents that build apps. CodeSandbox is a cloud development platform offering instant dev environments, micro-VMs (Codeflow), and AI features including SDK Agent for autonomous app building.

Open-source AI coding assistant for VS Code and JetBrains with chat, edit, autocomplete, and customizable agents. Continue is an open-source AI coding assistant for VS Code and JetBrains IDEs offering chat, inline edits, AI autocomplete, and custom agents.

Autonomous AI software engineer from UK startup Cosine, trained on real engineer behavior and notable for top SWE-bench Verified scores. Genie is an autonomous AI software engineer from Cosine, trained on traces of real human engineers solving tasks rather than purely on code.

Charm's beautiful, open-source terminal-based AI coding agent built in Go, with multi-model support and a polished TUI. Crush is an open-source AI coding agent from Charm, the makers of popular Go terminal libraries (Bubble Tea, Lip Gloss).

AI-first code editor (VS Code fork) with deep agent mode, multi-file edits, and tab-completion powered by frontier models.

Open-source dev environment management platform with cloud sandbox infrastructure for AI agents and developers. Daytona is an open-source platform for managing development environments, with a focus on providing fast, secure cloud sandboxes for AI agents and human developers.

Family of open-weight code-specialized LLMs from DeepSeek with strong performance on coding benchmarks at low inference cost. DeepSeek Coder is a family of open-weight code LLMs from DeepSeek, available in various sizes for self-hosting and via DeepSeek's API.

Open-source AI software engineer modeled on Devin, capable of breaking down tasks, browsing the web, and writing code with multiple LLMs. Devika is an open-source AI software engineer created as a community alternative to Devin.

Autonomous AI software engineer from Cognition that handles complete development tasks (coding, testing, deploying) via Slack and web.

Open-source multi-agent system from KuafuAI that automates software development by combining LLMs with traditional DevOps tooling and pipelines.

Mistral's open-weight coding model series optimized for agentic software engineering, designed to be the LLM backbone of OSS coding agents.

Open-source, local-first AI app builder positioned as a privacy-respecting alternative to Lovable and Bolt, running on your own machine.