# Buda > Buda is a drive-based AI agent platform that lets you build and manage entire AI-powered companies with persistent memory, browser automation, terminal access, and multi-agent orchestration. Buda is a cloud-based AI agent platform built around persistent, drive-based workspaces — enabling individuals and teams to deploy fleets of AI agents that remember context, execute tasks, and evolve over time. Unlike chat-based AI tools, Buda agents retain memory across sessions through a persistent cloud drive, making them capable of running long-term workflows across sales, marketing, coding, and operations. The platform features a unified visual workspace combining Agent, Drive, Browser, Terminal, and Git views so users can see every step their AI takes in real time. Buda positions itself as a human-led alternative to raw agent frameworks, keeping people in the director role while AI handles execution. - **Persistent Drive-Based Memory**: Every agent writes to a dedicated cloud workspace — files, context, and task history survive restarts so agents never start from zero. - **All-in-One Visual UI**: Agent, Drive, Browser, Terminal, and Git are accessible from a single dashboard, giving full visibility into agent actions. - **Multi-Agent Orchestration**: Run parallel agents simultaneously — coder, analyst, operator — all managed from one workspace with isolated sandboxes per agent. - **Browser Automation**: Agents can control a browser to fill forms, scrape leads, and interact with web apps autonomously. - **Scheduled Automations**: Set recurring tasks (daily reports, CRM syncs, PR reviews) that agents execute on a schedule without manual triggers. - **Agent Skills**: Equip agents with modular skills (Web Scraper, Data Analyst, GitHub integration) without writing code, via a Skills Manager UI. - **Multi-Platform Deployment**: Deploy agents to Discord, Telegram, Slack, WhatsApp, WeChat, Feishu/Lark, Teams, and Line for real-time team assistance. - **Multimodal File Understanding**: Agents read PDF, DOCX, PPTX, MP4, MP3, PNG, JPG, and more, with OCR support and instant search across binary files. - **Secure Isolated Sandboxes**: Each agent runs in an isolated drive sandbox powered by Sandock.ai, with permission controls and no cross-workspace memory leakage. - **Context Compression**: Smart pruning reduces token spend by up to 3× compared to raw API calls. - **Team & Enterprise Features**: Role-based access, SSO, audit logs, and self-hosted deployment options for larger organizations. - **Multiple Frontier AI Models**: Access to 12 models including Claude Sonnet 4.6, GPT-5.4, Gemini 3.1 Pro, and more — pay per credit consumed. ## Features - Persistent drive-based agent memory - Multi-agent orchestration - Browser automation - Terminal access - Git integration - Scheduled automations - Agent skills marketplace - Multi-platform deployment (Discord, Slack, Telegram, etc.) - Multimodal file understanding (PDF, DOCX, MP4, etc.) - OCR support - Isolated agent sandboxes - Context compression (3x token savings) - Role-based access control - SSO support - Audit logs - Self-hosted option - BYOK (Bring Your Own Key) - 12 frontier AI models - Real-time backup and two-way sync - Interactive visual workspace ## Integrations Discord, Telegram, Slack, WhatsApp, WeChat, Feishu/Lark, Microsoft Teams, Line, GitHub, Salesforce, Stripe, Sandock.ai, Claude (Anthropic), GPT (OpenAI), Gemini (Google), DeepSeek, MiniMax, Kimi ## Platforms MACOS, WEB, API, CLI ## Pricing Freemium — Free tier available with paid upgrades ## Links - Website: https://buda.im - Documentation: https://buda.im/docs - Repository: https://github.com/mr-kelly/far - EveryDev.ai: https://www.everydev.ai/tools/buda-im