Metis Agent, open source, that boosts coding performance by 50%

Metis, an open-source coding agent released by developer Wholiver, claims to boost coding performance by 50%, solving 73 of 89 tasks (82.02%) on Terminal-Bench 2.1 using DeepSeek V4 Flash, compared to OpenCode's 60 tasks (67.42%), a 14.6% improvement. The agent features recursive multi-agent delegation, SQLite memory, and verification gates, and is available under an MIT license with a free desktop app and CLI.

English · 简体中文 /Wholiver/metis/blob/main/README.zh-CN.md A coding agent that searches, remembers, executes, and verifies across terminal and desktop. Quick start quick-start · Benchmark & Comparison benchmark--comparison · Key features key-features · Documentation documentation Standalone application with built-in Metis CLI and Server runtime no Node.js required : macOS Apple Silicon : Download Metis- -macos-arm64.dmg from the latest GitHub Release https://github.com/Wholiver/metis/releases/latest and drag to Applications . Windows x64 : Download Metis- -win-x64-setup.exe or .zip from the latest GitHub Release https://github.com/Wholiver/metis/releases/latest . CLI installation Requires Node.js =22.19.0 . npm install -g @wholiver hu/metis metis Run in any repository or directory: metis "Explain this repository" metis @src/main.ts "Review this file" git diff | metis -p "Review this diff" Use /login for subscription providers or configure an API key. See Quickstart /Wholiver/metis/blob/main/docs/quickstart.md for the complete guide. In a controlled benchmark run using the same model DeepSeek V4 Flash , same 89 real-world tasks, identical budget, and environment: | Agent Framework | Model | Benchmark | Solved Accuracy | Architecture & Harness Advantage | |---|---|---|---|---| 🏆 Metis | DeepSeek V4 Flash | Terminal-Bench 2.1 89 tasks | 73 / 89 82.02% | ✅ Recursive 5-role agents + SQLite memory + Plan/Build | OpenCode | DeepSeek V4 Flash | Terminal-Bench 2.1 89 tasks | 60 / 89 67.42% | | 📈 Improvement | Same Model & Budget | Same Environment | +14.6% +13 tasks | 🚀 Harness, memory, and verification gates alone | | Capability | Metis | Claude Code | OpenCode | Cursor / Cline | |---|---|---|---|---| License & Pricing | ✅ MIT $0 Free | ❌ Proprietary | ✅ MIT $0 Free | | Model Freedom | ✅ Any Model / OrcaRouter | ❌ Anthropic Only | ✅ Multi-Provider | | User Interfaces | ✅ TUI + React Desktop | ||| Workflow Mode | ✅ Plan ↔ Build Dual-Mode | ||| Multi-Agent System | ✅ Recursive L0→L4 5 Roles | ❌ None | || Durable Memory | ✅ SQLite + Vector Search | ❌ Ephemeral | ❌ Ephemeral | | Verification Gates | ✅ Test Gates + Video Evidence | ||| Headless Benchmark | ✅ Python Adapter + Trace | ❌ None | ❌ None | Plan & Build Dual Workflows — Safely investigate in read-only Plan mode, then execute approved plans with a live-updating checklist in Build mode. Dual Interface for Terminal & Desktop — Work directly in your terminal via the rich interactive TUI, or use the dedicated React/Vite Desktop workspace on macOS and Windows. Recursive Multi-Agent System — Native named agents coordinator , planner , implementer , reviewer , verifier with L0→L4 recursive delegation and Git Worktree isolation. Durable Memory & Resumable Sessions — Project knowledge and decisions persist in SQLite across restarts, context compactions, and session forks. Extensible & Model-Agnostic — Use any LLM provider OpenAI, Anthropic, DeepSeek, OrcaRouter, Gemini, Groq, Ollama, vLLM and extend with TypeScript plugins, Agent Skills, and MCP. Benchmark-Grade Reliability — Automated verification gates, video evidence inspection, and full Terminal-Bench & Harbor evaluation readiness. | Topic | Guide | |---|---| | Install, authenticate, and start | | Using Metis /Wholiver/metis/blob/main/docs/usage.md · TUI /Wholiver/metis/blob/main/docs/tui.md Providers /Wholiver/metis/blob/main/docs/providers.md · Custom models /Wholiver/metis/blob/main/docs/models.md · Custom providers /Wholiver/metis/blob/main/docs/custom-provider.md Named Agents & Delegation /Wholiver/metis/blob/main/docs/agents.md TerminalBench & Harbor /Wholiver/metis/blob/main/docs/terminalbench.md Sessions /Wholiver/metis/blob/main/docs/sessions.md · Compaction /Wholiver/metis/blob/main/docs/compaction.md Extensions /Wholiver/metis/blob/main/docs/extensions.md · Skills /Wholiver/metis/blob/main/docs/skills.md · Packages /Wholiver/metis/blob/main/docs/packages.md Prompt templates /Wholiver/metis/blob/main/docs/prompt-templates.md · Themes /Wholiver/metis/blob/main/docs/themes.md · Keybindings /Wholiver/metis/blob/main/docs/keybindings.md SDK /Wholiver/metis/blob/main/docs/sdk.md · RPC /Wholiver/metis/blob/main/docs/rpc.md · JSON /Wholiver/metis/blob/main/docs/json.md Video tool /Wholiver/metis/blob/main/docs/video.md Security /Wholiver/metis/blob/main/docs/security.md · Settings /Wholiver/metis/blob/main/docs/settings.md Windows /Wholiver/metis/blob/main/docs/windows.md · Termux /Wholiver/metis/blob/main/docs/termux.md · tmux /Wholiver/metis/blob/main/docs/tmux.md · Containers /Wholiver/metis/blob/main/docs/containerization.md See the documentation index /Wholiver/metis/blob/main/docs/index.md for every guide. Developer information npm run build Compile TypeScript and copy runtime assets npm test Run the Vitest suite npm run clean Remove compiled output npm run build:binary Build the standalone binary npm --prefix desktop run dev Start the React/Vite Desktop app in development npm --prefix desktop run build Build the renderer and Electron artifact The package exports the Node.js SDK from @wholiver hu/metis and the RPC entry point from @wholiver hu/metis/rpc-entry . Contributions are welcome. See CONTRIBUTING.md /Wholiver/metis/blob/main/CONTRIBUTING.md for development, Extension and Package integration, testing, and AI-assisted contribution guidance. Distributed under the MIT License /Wholiver/metis/blob/main/LICENSE .