# Best new AI coding agents (October 2026)

Canonical: https://theagenticweekly.com/best/ai-coding-agents · Last updated: 2026-10-25 · 20 tools ranked

> Our current #1 coding agent is Claude Code, tested in a smoke test on 2026-10-25 with a local model rather than Claude models. That rank is an editorial pick, not a benchmark. Pick Cursor or GitHub Copilot if you want an editor, Codex CLI, OpenCode, Cline or Goose if you want open-source software, and read the price column before you commit: several prices are marked not verified.

Coding agents read your repo, edit files and run commands. They ship updates almost daily, so we only call out the releases that change which one you should use.

## Overall best AI coding agent

**Short answer:** our overall pick is [Claude Code](https://theagenticweekly.com/tools/claude-code), an editorial #1 rather than a benchmark result. In our smoke test on 2026-10-25 it [partly worked](https://theagenticweekly.com/tests/claude-code) with a small local model (not Claude models), so the test says little about how it performs with its vendor's own models.

- **Best overall (terminal):** [Claude Code](https://theagenticweekly.com/tools/claude-code), our editorial #1. It needs a paid Claude plan and is not on the Free plan.
- **Best for IDE-based daily work:** [Cursor](https://theagenticweekly.com/tools/cursor) and [GitHub Copilot](https://theagenticweekly.com/tools/github-copilot) remain practical choices.
- **Best open-source, model-flexible terminal agents:** [OpenCode](https://theagenticweekly.com/tools/opencode), [Cline](https://theagenticweekly.com/tools/cline), [Goose](https://theagenticweekly.com/tools/goose) and Qwen Code. OpenCode and Goose each [partly worked](https://theagenticweekly.com/tests/opencode) in [our smoke tests](https://theagenticweekly.com/tests/goose) on a small local model.
- **Which model to run inside an agent:** see [best LLM for coding](https://theagenticweekly.com/best/llm-for-coding) for sourced benchmark numbers with dates.

## AI coding agents compared: price, free tier and best use

Table: AI coding agents compared by price, free tier, best use and evidence label. Prices read from vendor pages on 2026-10-25 unless marked not verified.

| Tool | Price (read 2026-10-25) | Free tier | Best for | Evidence |
|---|---|---|---|---|
| [Claude Code](https://theagenticweekly.com/tools/claude-code) | Claude Pro $20/mo ($17/mo billed annually); Max from $100/mo | No: not on the Free plan | Terminal and VS Code work where the agent runs commands and edits files | Tested (smoke test) |
| [Codex CLI](https://theagenticweekly.com/tools/codex-cli) | Not verified (open-source software, Apache-2.0) | Not verified | Open-source terminal agent from OpenAI with frequent releases | Tested (smoke test) |
| [Gemini CLI](https://theagenticweekly.com/tools/gemini-cli) | Open source (Apache-2.0); paid API key or Code Assist licence for access | Unpaid-tier and Google One users are directed to Antigravity CLI (README, since 2026-06-18) | Enterprise and paid-API-key users who want to read and extend the code | Reported |
| [Cursor](https://theagenticweekly.com/tools/cursor) | Free $0; paid plans from $8/mo; Teams from $30/user/mo | Yes: Free plan with limited agent and chat requests | Working inside an editor rather than a terminal | Reported |
| [GitHub Copilot](https://theagenticweekly.com/tools/github-copilot) | Free $0; Pro $10/mo; Pro+ $39/mo; Max $100/mo | Yes: limited chat and agent usage, 2,000 completions a month; Free supports CLI and agent mode | Teams already on GitHub; IDE, CLI and GitHub.com in one | Reported |
| [OpenCode](https://theagenticweekly.com/tools/opencode) | Free software (MIT); you pay the model provider you connect | Software is free; model costs not verified | Switching between model providers; terminal UI plus desktop app | Tested (smoke test) |
| [Cline](https://theagenticweekly.com/tools/cline) | Free software (Apache-2.0); you bring your own model API keys | Software is free; model costs not verified | IDE users who want an open-source agent, with CLI and SDK | Reported |
| [Goose](https://theagenticweekly.com/tools/goose) | Free software (Apache-2.0); you choose the LLM | Software is free; model costs not verified | Desktop app plus CLI with subagent controls | Tested (smoke test) |
| [Antigravity CLI](https://antigravity.google/docs/cli/overview/) | Not verified (not on the pages we read) | Not verified | Google's terminal agent, sharing its agent harness with Antigravity 2.0 | Reported (not yet a full entry) |
| [Amp](https://ampcode.com/pricing) | Hobby free (pay as you go for hosted orbs); Individual $20/mo with 45,000 orb minutes | Yes: Hobby plan; self-hosted orbs and bring-your-own-keys are free of Amp fees | Agents that keep running in remote 'orbs' after you close your laptop | Reported (not yet a full entry) |
| [Qwen Code](https://github.com/QwenLM/qwen-code) | Free software (Apache-2.0); needs a model API key or plan | Software is free; model costs not verified | Open-source terminal agent that works with OpenAI, Anthropic, Gemini and Qwen APIs or local models | Reported (not yet a full entry) |
| [Kimi Code](https://www.kimi.com/code/docs/en/) | Kimi membership subscription (price not read) | Not verified | Terminal, desktop or VS Code agent included with a Kimi membership | Reported (not yet a full entry) |
| [Devin CLI](https://docs.devin.ai/cli) | Not verified | Not verified | Local terminal agent with an optional hand-off to Devin Cloud | Reported (not yet a full entry) |
| [Pi](https://github.com/badlogic/pi-mono/blob/main/packages/coding-agent/README.md) | Not verified | Not verified | A minimal, extensible terminal agent with TypeScript extensions | Reported (not yet a full entry) |

Prices change and several vendors bill by usage or credits, so a monthly price is not the whole cost. "Not verified" means we did not find the figure on an official page we read, not that none exists. Tools marked "not yet a full entry" have no Agentic Weekly write-up yet; their rows come from the linked official pages only and nothing about them was run. Evidence labels: see [methodology](https://theagenticweekly.com/methodology).

## Best picks by use case

These are editorial picks based on what each vendor says and, for Tested entries, the small smoke test we ran. They are not benchmark results.

- **Terminal-first, most actively updated:** [Claude Code](https://theagenticweekly.com/tools/claude-code), our #1. It needs a paid Claude plan (it is not on Free) and our test used a local model, not Claude models. See [Claude Code vs Cursor](https://theagenticweekly.com/compare/claude-code-vs-cursor).
- **You live in an editor:** [Cursor](https://theagenticweekly.com/tools/cursor), or [GitHub Copilot](https://theagenticweekly.com/tools/github-copilot) if your team already works in GitHub. Copilot's Free plan supports CLI and agent mode with limited usage, per its plans page.
- **You want open-source software you can read:** [Codex CLI](https://theagenticweekly.com/tools/codex-cli), [OpenCode](https://theagenticweekly.com/tools/opencode), [Cline](https://theagenticweekly.com/tools/cline), [Goose](https://theagenticweekly.com/tools/goose) and Qwen Code are all open source; you pay for whichever model you connect.
- **You want to switch model providers freely:** OpenCode, Cline and Qwen Code advertise support for many providers; Qwen Code lists OpenAI, Anthropic, Gemini and Qwen APIs plus local models.
- **You want agents to keep working while you are away:** Amp runs each thread in a remote machine it calls an orb, and its docs say the agent keeps working after you close your laptop.
- **You are on Google's stack:** Antigravity CLI is Google's terminal agent. Gemini CLI's README says unpaid-tier and Google One users moved to Antigravity CLI on 2026-06-18, so check which one applies to your account.
- **You want to spend nothing to try one:** Cursor and GitHub Copilot both list free plans with limits; Amp has a free Hobby tier; the open-source tools cost nothing but still need a model.

Rankings are editorial. Paid placements never affect them. Reported = written from official sources; Tested = a small smoke test whose scope is stated per tool (not a benchmark). How ranks are chosen: https://theagenticweekly.com/editorial-policy#ranking.

## Our ranked list

| Rank | Tool | Why | Verdict | Evidence | Last verified |
|---|---|---|---|---|---|
| 1 | [Claude Code](https://theagenticweekly.com/tools/claude-code) | One of the most actively maintained coding agents, with daily releases and detailed changelogs. | Update now if you use the sandbox or run Claude Code on a cloud provider; skip if you wanted flashy new features, this is a hardening release. | Tested (smoke test, 2026-10-25, local Qwen2.5-3B-Instruct, not vendor models) | 2026-10-25 |
| 2 | [Codex CLI](https://theagenticweekly.com/tools/codex-cli) | A mature open-source terminal agent that ships frequent updates; this one is about workflow polish. | A quality-of-life release: worth updating if you re-type the same prompts or switch models a lot, not a reason to switch tools. | Tested (smoke test, 2026-10-25, local Qwen2.5-3B-Instruct, not vendor models) | 2026-10-25 |
| 3 | [Gemini CLI](https://theagenticweekly.com/tools/gemini-cli) | Still maintained for enterprise and paid API key users, but Google steers everyone else to Antigravity CLI. | Skip it as a new starter on a consumer plan; keep it only if you are on an enterprise license or a paid API key. | Reported (not run by us) | 2026-10-25 |
| 4 | [Cursor](https://theagenticweekly.com/tools/cursor) | A full editor plus cloud agents; the October release focuses on multi-repo cloud environments. | Worth a look if you run agents on several repos or want one place for editor, cloud agents and Git hosting; the October changes are mostly for teams. | Reported (not run by us) | 2026-10-25 |
| 5 | [GitHub Copilot](https://theagenticweekly.com/tools/github-copilot) | Broad surface area (IDE, CLI, GitHub.com, agent app) and a rotating model list. | A sensible default if your team already lives on GitHub; check which plan you need for new models. | Reported (not run by us) | 2026-10-25 |
| 6 | [opencode](https://theagenticweekly.com/tools/opencode) | Open-source (MIT) coding agent with a large community; the latest release is a security-flavored runtime bump. | Update when convenient for the runtime security fixes; pick opencode if you want an open-source agent that is not tied to one model vendor. | Tested (smoke test, 2026-10-25, local Qwen2.5-3B-Instruct, not vendor models) | 2026-10-25 |
| 7 | [Cline](https://theagenticweekly.com/tools/cline) | Open-source agent with extension, CLI and SDK surfaces and very frequent releases. | A good open-source pick if you want an agent inside your IDE or a CLI and an SDK to embed it; the October notes are incremental. | Reported (not run by us) | 2026-10-25 |
| 8 | [goose](https://theagenticweekly.com/tools/goose) | Open-source agent that works with many LLMs and now has resumable subagents. | Pick goose if you want an open-source agent with real subagent controls; v1.56.0 is a solid, subagent-heavy update. | Tested (smoke test, 2026-10-25, local Qwen2.5-3B-Instruct, not vendor models) | 2026-10-25 |
| 9 | [Amazon Q Developer](https://theagenticweekly.com/tools/amazon-q-developer) | AWS coding assistant with IDE plugins and a CLI; plugins being retired. | AWS's AI coding assistant for IDEs and the CLI, with a free tier and a $19 Pro tier; IDE plugin support ends April 30, 2027. | Reported (not run by us) | 2026-10-25 |
| 10 | [Augment Code](https://theagenticweekly.com/tools/augment-code) | Context-engine coding agent and platform with flat team pricing. | Coding agent and 'Cosmos' platform with a Context Engine, CLI and cloud agents; plans are flat team prices with usage included. | Reported (not run by us) | 2026-10-25 |
| 11 | [Devin](https://theagenticweekly.com/tools/devin) | Cloud agent that runs in its own environment and opens pull requests. | Cognition's autonomous software engineer: a cloud coding agent that tests in its own browser and aims to hand back a mergeable pull request. | Reported (not run by us) | 2026-10-25 |
| 12 | [Kiro](https://theagenticweekly.com/tools/kiro) | Spec-driven agentic IDE and CLI. | AWS's agentic coding tool that turns prompts into requirements, designs and tasks (spec mode), with an IDE and a CLI. | Reported (not run by us) | 2026-10-25 |
| 13 | [Qodo](https://theagenticweekly.com/tools/qodo) | AI code review and quality tool for IDEs, PRs and security. | AI code review platform with 15+ agentic workflows for IDEs, pull requests and security, priced by team credits. | Reported (not run by us) | 2026-10-25 |
| 14 | [Warp](https://theagenticweekly.com/tools/warp) | Agentic terminal and cloud-agent platform with a free tier. | Warp's terminal and agent platform; it now pitches 'open infrastructure for cloud software factories' defined as code, with any model. | Reported (not run by us) | 2026-10-25 |
| 15 | [Zed](https://theagenticweekly.com/tools/zed) | Fast open-source editor with built-in agents and edit prediction. | High-performance, multiplayer code editor from the creators of Atom and Tree-sitter, open source with AI optional. | Reported (not run by us) | 2026-10-25 |
| 16 | [Kilo Code](https://theagenticweekly.com/tools/kilo-code) | Open-source agent for VS Code, JetBrains, CLI and cloud with 500+ models. | Open-source AI coding agent for VS Code, JetBrains, CLI and Cloud, with 500+ models at zero markup or your own keys. | Reported (not run by us) | 2026-10-25 |
| 17 | [OpenHands](https://theagenticweekly.com/tools/openhands) | Open-source coding agent platform with local and cloud use. | Open-source AI coding agent platform you can start locally, automate with workflows and run with any model. | Reported (not run by us) | 2026-10-25 |
| 18 | [Continue](https://theagenticweekly.com/tools/continue) | Apache-2.0 coding assistant with a CLI that worked against a local model. | A configurable open-source assistant; the CLI worked with a local model through a YAML config. | Tested (smoke test, 2026-10-25, local 3B model (Qwen2.5-3B)) | 2026-10-25 |
| 19 | [Windsurf (now Devin Desktop)](https://theagenticweekly.com/tools/windsurf) | AI IDE with an agent command center for local and cloud agents. | Windsurf was renamed Devin Desktop on June 2, 2026: a full IDE with an Agent Command Center for managing local and cloud coding agents. | Reported (not run by us) | 2026-10-25 |
| 20 | [Aider](https://theagenticweekly.com/tools/aider) | Open-source terminal pair programmer that works with almost any LLM. | Open-source AI pair programming in your terminal that maps your codebase and connects to almost any LLM, including local models. | Tested (smoke test, 2026-10-25, local 3B model (Qwen2.5-3B)) | 2026-10-25 |

## New and notable coding agents we have read about but not yet written up

These are real products with official documentation we read on 2026-10-25. We have not run any of them, and we do not rank them. Each will get a full entry only when we can say what we tested.

- **[Antigravity CLI](https://antigravity.google/docs/cli/overview/)** (`agy`): Google's keyboard-driven terminal agent. Its docs say it shares an agent harness and settings with Antigravity 2.0 and installs with a one-line script on macOS and Linux. It also runs non-interactively with `agy -p`.
- **[GitHub Copilot CLI](https://docs.github.com/en/copilot/get-started/cli-quickstart)**: installs with `npm install -g @github/copilot` (Node.js 22 or later) and requires an active Copilot subscription unless you bring your own model provider keys, per GitHub's docs. It is part of the Copilot entry above.
- **[Amp](https://ampcode.com/docs)**: a coding agent and development environment that runs each thread in a remote orb. It works from the web, a terminal CLI and mobile apps, and its pricing page lists a free Hobby tier and a $20/month Individual tier.
- **[Qwen Code](https://github.com/QwenLM/qwen-code)**: an Apache-2.0 terminal agent from the Qwen team. It installs with npm or Homebrew and you run `/auth` to configure a provider.
- **[Kimi Code](https://www.kimi.com/code/docs/en/)**: Moonshot AI's coding service, offered with a Kimi membership as a desktop app, a CLI and a VS Code extension. Its docs say the CLI was rebuilt on Node.js.
- **[Devin CLI](https://docs.devin.ai/cli)**: Cognition's local terminal agent with optional hand-off to Devin Cloud, installed with a one-line script.
- **[Pi](https://github.com/badlogic/pi-mono/blob/main/packages/coding-agent/README.md)**: a deliberately minimal terminal agent whose README describes interactive, print and JSON, RPC and SDK modes and TypeScript extensions.

## How to choose a coding agent

Start with where you work. If you spend your day in a terminal, the terminal agents (Claude Code, Codex CLI, OpenCode, Goose, Qwen Code, Antigravity CLI) fit naturally. If you want the agent next to your files, an editor product such as Cursor, or an extension such as Cline or Copilot, is a smaller change.

Next, decide who pays for the model. Claude Code, Cursor, Copilot and Amp bundle model usage into a plan, so you pay the vendor and accept its limits. The open-source agents are free to install but send requests to a model you pay for, or run a local one. Local models are cheaper but, in our own smoke tests with a small local model, were far weaker than the hosted models these tools are designed for.

Finally, check how fast the tool changes. These agents release almost daily, so a pick today may be stale next month. Each tool page shows the release we covered and the date we last verified it, and the [weekly archive](https://theagenticweekly.com/archive) shows what changed.

## How these rankings work

The ranked list above is editorial judgement on the tools we cover. It is not a benchmark, a popularity count or a paid position. Paid placements are labelled and never change a rank. The full basis is on the [methodology page](https://theagenticweekly.com/editorial-policy#ranking). A Tested label means a small smoke test with its scope stated: our tests ran on a Linux machine with no GPU, using a small local model, so they say little about how a tool performs with its vendor's own models.

## FAQ

### What is the best AI coding agent right now?

Our editorial #1 is Claude Code, based on how actively it is maintained and a smoke test that installed it and ran a short task. The test used a local model, not Claude models, so it is not a quality benchmark. Cursor and GitHub Copilot are our picks if you prefer an editor.

### Which AI coding agents have a free tier?

Per their pricing pages read on 2026-10-25, Cursor and GitHub Copilot list free plans with limited usage, and Amp lists a free Hobby tier. Claude Code is not on Claude's Free plan. Open-source agents such as OpenCode, Cline, Goose and Qwen Code are free to install but need a model you pay for or run locally.

### Which AI coding agents are open source?

Codex CLI, Gemini CLI, Goose, Cline, OpenCode and Qwen Code are open source according to their repositories. Claude Code, Cursor and GitHub Copilot are vendor products.

### Did you test all of these agents?

No. Claude Code, Codex CLI, OpenCode and Goose are Tested as smoke tests with the scope stated on each tool page. The others are Reported from official sources. Newly listed agents such as Amp, Qwen Code, Kimi Code, Devin CLI, Pi and Antigravity CLI were not run.

### Is Gemini CLI still the way to use Gemini in the terminal?

Its README says Gemini CLI was replaced by Antigravity CLI for unpaid-tier and Google One users on 2026-06-18, and that Gemini CLI is not optimized for the latest Gemini models. Check the README for your account type.

## Sources

- [Claude pricing](https://claude.com/pricing) (read 2026-10-25)
- [Cursor pricing](https://cursor.com/pricing) (read 2026-10-25)
- [GitHub Copilot plans](https://github.com/features/copilot/plans) (read 2026-10-25)
- [GitHub Copilot CLI quickstart](https://docs.github.com/en/copilot/get-started/cli-quickstart) (read 2026-10-25)
- [Gemini CLI README](https://github.com/google-gemini/gemini-cli) (read 2026-10-25)
- [Antigravity CLI overview](https://www.antigravity.google/docs/cli/overview/) (read 2026-10-25)
- [Amp pricing](https://ampcode.com/pricing) (read 2026-10-25)
- [Qwen Code repository](https://github.com/QwenLM/qwen-code) (read 2026-10-25)
- [Kimi Code docs](https://www.kimi.com/code/docs/en/) (read 2026-10-25)
- [Devin CLI docs](https://docs.devin.ai/cli) (read 2026-10-25)
- [Pi coding agent README](https://github.com/badlogic/pi-mono/blob/main/packages/coding-agent/README.md) (read 2026-10-25)

## How to cite

Agentic Weekly, "Best AI coding agents (updated weekly)", last updated 2026-10-25, https://theagenticweekly.com/best/ai-coding-agents