Coding agents
Claude Code
Anthropic's agentic coding tool: understands your codebase, edits files and runs commands.
Strongest at multi-file changes in real repos. Lives in the terminal, so it fits any editor.
Verified
Loading…
AI tool directory / Coding agents / Alternatives
Muse Code: Meta's multi-agent coding agent for Mac and Windows, built for Muse models, with parallel workers and background reviewers. These 24 coding agents do similar jobs, listed from highest to lowest RobotWise score. Descriptions come from each vendor's own site.
Anthropic's agentic coding tool: understands your codebase, edits files and runs commands.
Strongest at multi-file changes in real repos. Lives in the terminal, so it fits any editor.
Verified
AI code editor where you hand off tasks to agents that turn ideas into code.
Verified
Works from an open issue to a reviewed pull request, in your editor, terminal, on GitHub, and in JetBrains Air. In Air, Copilot runs local agent tasks and cloud agent tasks on a Copilot subscription.
Verified · Available in JetBrains Air for local IDE tasks and cloud agent tasks.
OpenAI's coding agent for writing, fixing and reviewing code.
Verified
Cognition's autonomous AI software engineer. SWE-2 Flash is in Devin Desktop and CLI, free through Nov 30, 2026.
Verified · SWE-2 Flash is in Devin Desktop and CLI, free through Nov 30, 2026.
Open-source AI agent that brings Gemini into your terminal.
Verified
Google's agent-first developer platform, with an app, IDE, CLI, extensions and SDK.
Verified
Agentic engineering tool that turns prompts into executable specs and builds across large codebases with parallel agents. Crew bots are generally available in Kiro Crew: named, long-running agents with their own memory and sessions. Announced November 5, 2026.
Verified · Crew bots are generally available in Kiro Crew.
Open-source coding agent for your editor that works with the model provider of your choice.
Verified
Independent AI code review for pull requests, with PR triage, change explanations and continuous security monitoring.
Verified
Model-independent coding agent. Droid 1.0 is the harness for complex projects and everyday development, and the Droid SDK runs that same agent from Node with multi-turn sessions and custom MCP tools.
Verified · Droid 1.0 harness and the Droid SDK.
Google's autonomous coding agent that works asynchronously on your GitHub repos.
Verified
Open-source AI coding agent for your terminal, IDE or desktop that works with models from any provider.
Verified
Coding agent and dev environment built for frontier models.
Verified
Warp Factories runs fleets of coding agents defined in code, triggered from Slack, Linear, Jira or GitHub, plus an agent CLI that works in any terminal.
Verified
AI pair programming in your terminal, editing code in your local git repo.
Verified
Ramp's collaborative coding agent that turns a task from Slack, GitHub, Linear, or the web into a reviewed pull request in a shared cloud session. Early access opened October 27, 2026.
Verified
Open-source platform for building and running AI coding agents: start locally, automate engineering workflows with any model, then scale with enterprise controls.
Verified
AI code review with deep codebase context and 15+ agentic workflows across IDEs, pull requests and security.
Verified
Open-source research prototype of a proactive coding agent for one GitHub repository. It listens to GitHub and optionally Feishu, then works a requested task in its own session. MIT license. Published October 31, 2026; not an official DeepSeek product.
Verified
Independent Git forge for agents and humans: Trails replace pull requests with agent-scored evals, with distributed Git hosting and Continuous Memory so agents can search past sessions. Announced October 26, 2026.
Verified
Open-source AI agent that runs on your machine as a desktop app, CLI or API, for code, workflows and everyday tasks.
Verified
Open-source coding-agent runtime in one Rust binary. It is model-agnostic and MCP-native, and runs as a CLI, a daemon, or an ACP agent in an editor. Apache-2.0.
Verified
Ricursive's open-weight dense model for agentic coding, with up to 1M context. Apache-2.0 weights are on Hugging Face.
Verified