✍️ AI Writing
See all writing tools →
💻 AI Coding
See all coding tools →
💻
Cursor
AI Coding
9.4
out of 10

The AI code editor that's changed how developers work. Codebase-wide context, multi-file edits, and natural language refactoring.

Code editorAutocompleteFree plan
🐙
GitHub Copilot
AI Coding
8.8
out of 10

The best value AI coding tool. Works in every major IDE at $10/month — no switching editors required.

Multi-IDEFree plan$10/mo
🌊
Windsurf
AI Coding
9.1
out of 10

A genuine Cursor rival with a more approachable UX. Cascade AI handles complex codebase-wide tasks impressively well.

AgenticFree planBeginner-friendly
Bolt.new
AI Coding
8.3
out of 10

Fastest path from idea to live app. Build full-stack MVPs from a single prompt directly in your browser.

No-codeFull-stackPrototyping
🔁
Replit
AI Coding
8.3
out of 10

The best browser-based AI coding environment. Replit Agent builds full apps from plain-language prompts with zero local setup required.

Browser-basedAI AgentFree plan
💗
Lovable
AI Coding
8.1
out of 10

The full-stack app builder leading the "vibe coding" category. One prompt generates a React + Supabase app and deploys it — but the credit system is opaque.

No-codeFull-stackSupabase
🛸
Google Antigravity 2.0
AI Coding
7.5
out of 10

Google's agent-first development platform. Parallel multi-agent execution and deep Google ecosystem integration — undermined by a rocky 2.0 launch and a CLI that wasn't shippable on day one.

Multi-agentGemini 3.5Free plan
🐋
DeepSeek V4
AI Coding
8.1
out of 10

Open-weight LLM scoring 80.6% on SWE-bench at $0.435/M input — roughly 1/30th the cost of Claude Opus 4.7. Frontier coding benchmarks at a developer-API price, with real data-residency caveats.

Open weights1M contextOpenAI-compatible
✳️
Claude Code
AI Coding
8.0
out of 10

Anthropic's coding agent — 1M context, Opus 4.7 at 87.6% SWE-bench, and Agent Teams running parallel instances on git worktrees. The strongest agent for senior developers; team pricing is the catch. Includes a full conflict-of-interest disclosure.

Agent Teams1M contextOpus 4.7
🌙
Kimi Code
AI Coding
7.8
out of 10

Moonshot AI's open-source Claude Code alternative. Kimi K2.6 scores 80.2% on SWE-bench at $0.60/M input — 25x cheaper than Claude Opus 4.7 — and runs every Claude Code MCP server unchanged. Agent Swarms coordinate up to 300 parallel agents. CLI-only, with the same data-residency caveats as DeepSeek.

Open weightsMCP compatibleAgent Swarms
🛠️
Grok Build
AI Coding
6.7
out of 10

xAI's terminal-native coding agent, and the third lab-backed entry alongside Claude Code and Codex CLI. Its 8 parallel sub-agents in isolated Git worktrees are a genuine architectural first — but grok-build-0.1 scores 70.8% SWE-Bench (vs 87.6% for Claude Code), Arena Mode isn't live yet, and $99/mo intro pricing reverts to $299/mo after 6 months. An early beta worth watching.

Git worktreesMCP compatibleEarly beta
🜂
MiniMax M3
AI Coding
7.9
out of 10

The first open-weight model to combine frontier coding (59% SWE-Bench Pro), a 1M-token context, and native text/image/video input. The new MSA sparse-attention architecture runs 15.6x faster at 1M context, and $0.60/M input is ~12x cheaper than Claude Opus 4.7. The catches: vendor-reported benchmarks awaiting independent verification, weights not shipped at launch, and the same data-residency questions as DeepSeek and Kimi.

Open weightsMultimodal1M context
🧠 AI Productivity
See all productivity tools →
🎨 AI Image
See category →
⚡ Automation
See category →
🎬 AI Video
See category →