Pro
$20/mo
Basic Claude Code access (legacy)
Agent of the Month
Claude Code
Claude's agentic coding mode, right in your terminal.
Anthropic's official agentic coding tool, available in the terminal, VS Code, JetBrains IDEs, and the Claude desktop, web, and mobile apps. Runs on Claude Sonnet 4.6, Opus 4.7, and Haiku 4.5 with extended thinking for deep multi-file reasoning.
+Pros
- Deepest Anthropic model integration — always first access to Claude Opus, Sonnet, and Haiku releases
- Subagents for true parallel task execution: e.g., backend and frontend agents running simultaneously
- Agent view (`claude agents`) gives a unified dashboard of all running/blocked/completed sessions
- Claude Agent SDK for embedding the same runtime into custom agentic applications
- Computer use in research preview — Claude can interact with native GUIs from the terminal
- Ultraplan: draft and review multi-step plans in a web editor before executing them remotely
- MCP tunnels, HTTP hooks, and /loop for external integrations and scheduled automation
−Cons
- Claude Code removed from the $20 Pro plan (Apr 2026) — now requires Max 5x ($100/mo) for new subscribers
- Terminal-first UX — no standalone IDE; requires comfort with CLI workflows
- Costs can escalate quickly on API usage plans for heavy or parallel workloads
- Opus 4 and 4.1 deprecated (May 2026) — model continuity less predictable than Cursor's multi-vendor approach
- Auto mode and computer use still in research preview — not production-ready
Pricing
Max
$100/mo
Full Claude Code access, extended thinking
Team Premium
$150/user/mo
Team features, Ultraplan
API
Metered
Pay-per-token usage
Claude Code is Anthropic's official agentic coding tool. It runs in the terminal as an npm package, in VS Code and JetBrains via official extensions, and inside the Claude desktop, web, and mobile apps — a single agent across every surface a developer touches.
Powered by Claude Sonnet 4.6, Opus 4.7, and Haiku 4.5, Claude Code uses extended thinking to reason through problems before writing code, producing higher-quality results on complex tasks. The agent autonomously reads files, edits code across many files in a single turn, runs terminal commands, reads test output, and iterates until the task is complete — with detailed visibility into every action.
Key Features
• Extended thinking: reasons through hard problems before coding
• Multi-file edits: coordinated changes across an entire project in one turn
• Full-codebase context: indexes the project, not just open files
• Native MCP: connects to filesystem, git, web, and any custom MCP server
• Multi-surface: terminal, VS Code, JetBrains, plus desktop and mobile apps
• Plan mode and skills: structured planning and reusable filesystem-based capabilities
• Visibility: every read, edit, and command is logged and reviewable
Version History
Agent view: `claude agents` shows all running, blocked, and completed sessions in one screen. `/goal` command keeps Claude working across turns until a condition is met. Fast mode defaults to Opus 4.7. Rewind menu can compress earlier context with 'Summarize up to here'. Plugins loadable from .zip archives and URLs.
Claude Opus 4.7 lands as new default on Max and Team Premium with `xhigh` effort level. Ultraplan enters early preview: draft cloud plans from CLI, review in web editor, run remotely or pull back local. Computer use in research preview (native app interaction from terminal). Auto mode research preview: AI classifier handles permission prompts.
Subagents GA for parallel task delegation (e.g., backend agent + frontend agent running concurrently). Checkpoints for safe rollbacks during autonomous runs. Claude Agent SDK (formerly Claude Code SDK) released with subagent orchestration and hooks support. Background daemons, HTTP hooks, and /loop command for scheduled automation.
Signature Snippet
$ claude 'Refactor the authentication middleware to use the new token cache'
Claude reads src/middleware/auth.ts, src/lib/token-cache.ts...
→ Moving token validation to the cache layer
→ Updating middleware to use new context
→ Writing tests for the refactored flow
▶ Applying changes to 3 files...
▶ All tests pass (12/12)More in this Space
Prime Agent (Prime Intellect)
Prime Agent is an open-source (MIT) self-improving coding and research harness by Prime Intellect, built on a Recursive Language Model that runs inside a persistent IPython REPL and a Continual Harness that refines its own state via /refine. It took the same frontier weights from 30.2% to 95.5% on ARC-AGI-3 RHAE Best@1, powered daemon-backed sessions, and ships no hosted offering.
OpenCodeReview (ocr)
OpenCodeReview (ocr) is Alibaba's open-source AI code-review CLI. It reads git diffs and drives an LLM agent through a hybrid architecture where deterministic pipelines handle file selection, rule matching, and comment positioning while the agent does the reasoning. Built on two years of internal use at Alibaba scale, it ships a multi-language ruleset and consumes roughly 1/9 of the tokens of general-purpose agents.
Delta (by Zed)
Multiplayer environment from Zed for coding with agents. Threads unify the agent conversation, the code edits, and human review in one shareable artifact, replacing the branch/PR loop. Built on DeltaDB, a CRDT-based version control layer that records every edit between commits alongside the prompts and reasoning that produced them.