AgentComparison
For: Developers

AI coding agents

Agents that write, change, test and ship code in your repository or in the cloud.

14 agentsVendor documentation, attributedNo hands-on tests

Who this group is for

Developers and teams who review code and want routine engineering work handed off.

What to check first: Repository write access, where code runs, who reviews changes, and usage or credit limits.

Agents in this group

product · coding agent

Claude Code

Anthropic's terminal coding agent for migrations, bug fixes and refactors.

Claude Code reads repositories, edits files, runs commands and connects development tools through terminal, IDE, desktop and web surfaces. Its current security docs describe Auto mode as the starting mode for interactive terminal and VS Code sessions, while Manual, Plan and other modes change approvals. Local execution does not mean code and prompts stay offline.

Documentation researched

Pricing: Pro $20/month includes Claude Code; Max $100-$200/month; API token billing option

Best fit: Developers delegating repository changes from a terminal.

product · software engineering agent

Devin

An autonomous software engineer with its own IDE, shell and browser.

Cognition documents Devin as an autonomous software engineer that writes, runs and tests code, available as Devin Cloud sessions with an IDE, shell and browser, plus Devin Desktop and a Devin CLI that can also steer cloud sessions from a terminal.

Documentation researched

Pricing: Free $0; Pro $20/month; Max $200/month; Teams $80/month minimum plus $40/month per full seat

Best fit: Engineering teams delegating well-scoped, verifiable tickets who will review the draft PRs it returns.

product · application building agent

Replit Agent

Builds and fixes whole apps from plain-language instructions inside Replit.

Replit documents an agent that creates applications, sets up projects, checks its work and fixes problems from natural-language instructions.

Documentation researched

Pricing: Core $20/month ($18 annual); Pro $100/month ($90 annual); monthly credits included; effort-based overage

Best fit: Building apps and related artifacts from natural-language requests.

product · coding agent (editor and cloud agents)

Cursor cloud agents

An AI code editor whose cloud agents work on branches and open pull requests.

Cursor is an AI code editor that also runs cloud agents: isolated virtual machines that clone your repository, work on a separate branch, test the change and push it for review. Agents can be started from the editor, web, mobile, Slack, GitHub, Linear or the API.

Documentation researched

Pricing: Free $0; Plus $8; Premium $30 (cheapest tier with cloud agents); Super $100; Ultra $200/month; Teams from $30/user (vendor pricing)

Best fit: Developers who want background work on repository branches and can review diffs, tests and artifacts before a human merges the draft pull request. Configure restricted egress and minimum repository access first.

product · coding agent (Google)

Jules

Google's background coding agent that turns GitHub tasks into pull requests.

Jules is Google's asynchronous GitHub coding agent. In its web workflow, you select a repository and branch, describe a task, approve a plan, then review the changes and publish a pull request. Code runs in a cloud VM. The API also supports automated workflows, but its approval defaults differ from the web app.

Documentation researched

Pricing: No-cost plan: 15 tasks per rolling 24 hours; higher limits via Google AI Pro / Ultra, local subscription prices unverified

Best fit: GitHub developers with a working project setup who want bounded bug fixes, dependency maintenance, documentation or tests returned for review while they work on something else.

product · spec-driven coding agent and IDE

Kiro

A coding agent that turns prompts into specs and tasks, then builds with parallel agents.

Kiro turns prompts into requirements, design and sequenced tasks, then implements them with parallel agents locally or in the cloud. Its site also describes property-based tests, a headless CLI for CI and model choice including Claude, GPT and open-weight models.

Documentation researched

Pricing: Free 50 credits; Pro $20/month for 1,000 credits; up to $200/month (vendor pricing)

Best fit: Developers who want specs and checks before the agent writes code.

product · coding agent on GitHub

GitHub Copilot cloud agent

GitHub's background coding agent that plans, edits a branch and opens pull requests.

GitHub's docs say Copilot cloud agent can research a repository, make an implementation plan and change code on a branch, then let you review the diff, iterate and create a pull request. It works in its own ephemeral environment powered by GitHub Actions, where it can run tests and linters.

Documentation researched

Pricing: Cloud agent: paid plans from $10/user/month; AI credits plus Actions minutes. Free plan does not include cloud agent.

Best fit: Teams already on GitHub that can split work into single-repository tasks, review generated changes and run checks before a human merges the pull request.

product · AI app builder

Lovable

Builds full web apps from plain-language prompts, with credits and a free daily grant.

Lovable describes a full-stack AI development platform for building, iterating on and deploying web applications from natural language, including frontend, backend, database and authentication. Work happens in shared workspaces that pool credits.

Documentation researched

Pricing: Free (daily build credits); Pro from $25/month (100 credits); Business from $50/month (100 credits) (vendor docs, read 2 Nov 2026)

Best fit: People who want a working web app from a description and are fine reviewing and owning the generated code.

product · agentic development platform

Google Antigravity

Google's agent-first dev platform: desktop app, CLI, IDE and SDK, with a free individual plan.

Google Antigravity is Google's agentic development platform. The vendor site describes four surfaces: a desktop app for running several local agents in parallel and grouping chats into Projects, a terminal CLI, an IDE with an agent manager and artifacts, and a Python SDK for building custom agents on the same harness.

Documentation researched

Pricing: Free individual plan with weekly limits; Google AI Pro price not stated (vendor pricing)

Best fit: Developers who want to try an agent-first IDE and CLI without paying, within weekly rate limits.

product · AI app and website builder

Bolt

Builds websites and apps from a chat prompt, with hosting and databases; priced by tokens.

Bolt builds websites, apps and prototypes from a chat prompt, and can import from Figma or GitHub. The vendor site describes a Bolt Agent that routes between models, tests and iterates on its own output, and Bolt Cloud hosting with databases, authentication and integrations.

Documentation researched

Pricing: Free (1M tokens/month, 300K daily); Pro $25/month; Teams $30/member/month (vendor pricing, read 2 Nov 2026)

Best fit: People who want a working web app or site from a prompt and are fine with token-based limits.

product · coding agent (CLI, IDE, web and desktop app)

OpenAI Codex

OpenAI's coding agent for terminal, editor, cloud and desktop, bundled with ChatGPT plans.

Codex is OpenAI's coding agent. OpenAI's docs describe a CLI that inspects code, edits files and runs commands in your repository, plus an IDE extension, cloud tasks and the ChatGPT desktop app. It is included in ChatGPT plans from Free to Pro and Business, or can be used with an API key at API prices.

Documentation researched

Pricing: Free $0, Go $8, Plus $20, Pro from $100 per month; Business $20/user billed annually

Best fit: Developers who already use or are willing to buy a ChatGPT plan and want one coding agent across terminal, editor and cloud, with sandbox and approval settings they control.

product · coding agent (editor and cloud agents)

Windsurf

Cognition's agentic coding editor with frontier models and Devin Cloud agents. Free tier; Pro $20/mo - paid tiers can opt out of training.

Windsurf is Cognition's agentic coding editor: Tab completions, inline edits and agent coding in the IDE, with frontier models (OpenAI, Claude, Gemini), Cognition's own SWE 1.7, and Devin Cloud agents on paid plans.

Documentation researched

Pricing: Free $0; Pro $20/mo; Max $200/mo; Team $80/mo + $40/dev seat

Best fit: Developers who want an agentic IDE with frontier-model choice and are comfortable opting out of training on a paid tier.

product · agentic terminal and coding agents

Warp

An agentic development terminal with cloud agents, codebase indexing and model choice. Free tier; Build $20/mo with $20 usage included.

Warp is an agentic development terminal: a modern terminal with Warp Agent CLI, cloud agents, codebase indexing and a model picker across OpenAI, Anthropic, Gemini, GLM and Kimi models, plus Warp Drive for team workflows.

Documentation researched

Pricing: Free $0; Build $20/mo; Max $200/mo; Business $50/user/mo

Best fit: Developers who live in the terminal and want agent runs, codebase indexing and model choice without leaving it.

product · coding agent prototype (self-hosted, DeepSeek Harness bundle)

DeepSeek dsh-superagent

DeepSeek's open-source prototype of a proactive coding agent for one GitHub repository, built as a DeepSeek Harness bundle.

dsh-superagent is a proactive coding agent for one GitHub repository, published 31 October 2026 as an open-source bundle for DeepSeek Harness (DSH). It listens to GitHub and, optionally, Feishu; when someone on an allow-list asks for work, it opens a work item, works in its own agent session and directory, and answers in the original conversation, for example with a pull request. A web panel shows what needs a person. The vendor is explicit that this is a research prototype for reference only, not an official DeepSeek product and not production-ready software.

Documentation researched

Pricing: Free (MIT); model provider costs depend on your DeepSeek Harness setup

Best fit: Teams already running DeepSeek Harness who want to study or adapt a working example of a proactive, low-attention coding agent for a single repository, and are comfortable operating research-grade self-hosted code.

To compare two agents side by side, use the comparison tool. Filter every agent by group and price in the directory.

Other groups