The Universal AI Agent for Your Terminal.
Connect to any major LLM provider through a single, unified terminal interface. 8 providers, 65+ commands, 22 built-in tools, smart model routing, multi-agent orchestration, and a full-featured TUI.
npm (all platforms):
npm install -g @openanalystinc/openanalyst-cliWindows (PowerShell):
irm https://raw.githubusercontent.com/OpenAnalystInc/cli/main/install.ps1 | iexmacOS / Linux:
curl -fsSL https://raw.githubusercontent.com/OpenAnalystInc/cli/main/install.sh | bashopenanalyst # launch the TUI
/login openai sk-... # authenticate inside the TUIOr set your API key before launching:
export OPENAI_API_KEY=sk-...
openanalyst| Feature | OpenAnalyst CLI |
|---|---|
| Providers | OpenAnalyst, Anthropic, OpenAI, Google Gemini, xAI, OpenRouter, Amazon Bedrock, Stability AI |
| Commands | 65+ slash commands — /commit, /pr, /model, /knowledge, /image, /bughunter, /swarm |
| Tools | 22 built-in — bash, PowerShell, file I/O, grep, glob, web search/fetch, REPL, agents, KB |
| TUI | Full terminal UI with branded banner, inline tool cards, diff view, streaming markdown |
| Model Routing | Smart per-task routing — fast models for explore, balanced for code, capable for planning |
| Agents | Autonomous agent, swarm execution, background tasks |
| Knowledge Base | Agentic RAG with vector + graph search |
| Voice | Microphone input with Whisper transcription |
| Sessions | Auto-save, resume, team sharing via git |
| MCP | Model Context Protocol with Playwright browser tools |
| Permissions | 4 modes — Default, Plan, Accept Edits, Danger |
| Cost Tracking | Per-model session costs with real-time display |
| API Server | HTTP API with SSE streaming for remote/cloud deployment |
| Team Collaboration | Sessions, todos, and plans shared via .openanalyst/ in git |
| Provider | Auth | Models |
|---|---|---|
| OpenAnalyst | API key | openanalyst-beta |
| Anthropic | API key | Opus 4.6, Sonnet 4.6, Haiku 4.5 |
| OpenAI | API key | GPT-5.4, GPT-4.1, GPT-4o, o3, o4 Mini, Codex Mini |
| Google Gemini | OAuth or API key | Gemini 3.1 Pro, Gemini 3 Flash, Gemini 2.5 Pro/Flash |
| xAI | API key | Grok 4, Grok 4 Fast, Grok 3 |
| OpenRouter | API key | 350+ models (DeepSeek V4, Llama 4, Mistral, etc.) |
| Amazon Bedrock | API key | Bedrock Claude |
| Stability AI | API key | Stability models |
Switch models mid-conversation with /model sonnet-4.6 — context persists across providers.
/help Show all commands and keybindings
/login Authenticate with any AI provider
/logout Remove credentials
/model Switch AI model
/models List all available models
/permissions Show or set permission mode
/commit Generate commit message and commit
/pr Create a pull request
/knowledge Search the knowledge base (Agentic RAG)
/image Generate an image
/bughunter Scan codebase for bugs
/diff-review AI-powered code review
/openanalyst Run autonomous agent loop
/swarm Spawn parallel agents
/compact Compact conversation context
/resume Resume a previous session
/vim Toggle vim mode
/sidebar Toggle sidebar panel
/exit Exit to terminal
65 commands total — type /help to see all with aliases.
| Key | Action |
|---|---|
| Enter | Submit prompt |
| Shift+Enter | New line |
| Ctrl+P | Cycle permission mode (Default → Plan → Accept Edits → Danger) |
| Ctrl+E | Toggle sidebar panel |
| Ctrl+C | Stop AI / quit (double-press to exit) |
| Ctrl+B | Run current prompt in background |
| Ctrl+L | Clear chat |
| Esc | Stop AI execution / enter scroll mode |
| Up/Down | Input history navigation |
| j/k | Navigate messages in scroll mode |
| y | Copy focused message to clipboard |
OpenAnalyst intelligently selects the right model for each task:
| Task Type | Model Tier | Example (Anthropic) | Example (OpenAI) |
|---|---|---|---|
| Explore (file search, scanning) | Fast | Haiku 4.5 | GPT-5.4 Nano |
| Research (reading, understanding) | Balanced | Sonnet 4.6 | GPT-5.4 Mini |
| Code (writing, editing) | Balanced | Sonnet 4.6 | GPT-5.4 Mini |
| Write (planning, complex reasoning) | Capable | Opus 4.6 | GPT-5.4 |
Change routing in the sidebar (Ctrl+E → navigate to Routing → Enter to cycle tier).
When your project has .openanalyst/ checked into git, team members get:
- Shared conversations — resume any team member's AI session
- Shared tasks — see pending todos and plans
- Shared instructions — project-specific OPENANALYST.md
- Shared settings — consistent model routing and permissions
Each developer uses their own API keys. Credentials are never shared.
Deploy OpenAnalyst on a remote server for API access:
openanalyst serve --port 8080# Simple one-shot query
curl -X POST http://localhost:8080/v1/chat \
-H "Content-Type: application/json" \
-d '{"message": "explain this codebase"}'
# Session-based with SSE streaming
curl -X POST http://localhost:8080/sessions
curl http://localhost:8080/sessions/session-1/eventsGlobal config: ~/.openanalyst/
Project config: .openanalyst/ (checked into git)
~/.openanalyst/
├── OPENANALYST.md # Global AI instructions
├── settings.json # Global settings
├── .env # API keys
├── credentials.json # Credential storage
├── commands/ # Custom slash commands
├── rules/ # Global rules
├── agents/ # Custom agent definitions
└── sessions/ # Chat history
- Email: support@openanalyst.com
- Website: openanalyst.com
- GitHub: OpenAnalystInc/openanalyst-cli
Developed by OpenAnalyst Inc. All rights reserved.