# Genii Personal open-source assistant with a local backend and web interface. The agent works through OpenClaw in an E2B sandbox; the published core is a development milestone. Canonical profile: https://agenten-kompass.mmeierhoff.chatgpt.site/en/agents/genii/ Editor: [Michael Meierhoff](https://agenten-kompass.mmeierhoff.chatgpt.site/en/about/) Sources reviewed: 2026-10-09 Basis: provider documentation. Examples are editorial suggestions, not hands-on product tests. ## Overview The public repository describes the local Genii core; OpenClaw runs the agent in an E2B sandbox. - Provider: Intelligent Internet - Product family: Genii / Intelligent Internet - Profile type: Local, open-source personal assistant (early development milestone) - Environment: Local backend and web interface; agent computer in an external E2B sandbox - Connections: E2B / OpenClaw according to the project - Setup: Technical setup; note milestone maturity - Pricing model: Genii software is Apache-2.0 licensed. OpenAI model access and the E2B agent computer are billed separately. ## Operation & oversight Operational details for this profile have not yet been researched. ## Provider Intelligent Internet publishes open AI tools and develops Genii as a personal assistant. The Genii repository and community documentation describe local development status and connected services. - Operator: Intelligent Internet - Organization: Project; operator's legal form not publicly stated - Website: https://ii.inc/ - Reviewed: 2026-10-09 - Repository: https://github.com/Intelligent-Internet/genii - GitHub-Stars: 27 (2026-10-09) - License: Apache-2.0 - Provider source: https://ii.inc/products - Provider source: https://github.com/Intelligent-Internet/genii - Provider source: https://api.github.com/repos/Intelligent-Internet/genii ## Prices and included usage Pricing checked: 2026-10-09 Genii software is Apache-2.0 licensed. OpenAI model access and the E2B agent computer are billed separately. ### Genii USD 0 software license Local development software; your own runtime and service accounts are required. Price source: https://github.com/Intelligent-Internet/genii ### E2B Hobby USD 0 base fee + usage USD 100 one-time credit; sandbox sessions up to one hour. Price source: https://e2b.dev/pricing ### E2B Pro USD 150/month + usage Sandbox sessions up to 24 hours; compute is billed by runtime. Price source: https://e2b.dev/pricing Self-hosting adds infrastructure and, where applicable, model costs. OpenAI's price list is one example of separately billed model access: https://developers.openai.com/api/docs/pricing Composio and optional Linq/tunnel services may add costs; they are separately configured integrations. Pricing details: https://github.com/Intelligent-Internet/genii ## Key features ### Local web interface Browser-based home page and API run locally. Source: https://github.com/Intelligent-Internet/genii ### Your own API key Uses its own OpenAI key, encrypted in the local backend. Source: https://github.com/Intelligent-Internet/genii ### Agent computer Uses an E2B desktop environment with a real browser and OpenClaw interface. Source: https://github.com/Intelligent-Internet/genii ### Persistent memory A plugin preserves context and a style profile shaped by feedback. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/memory.md ### Schedules and triggers Schedules and curated Gmail or Calendar events can activate Genii while the local backend service is running. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/server-trigger-classification.md ### Voice mode Talk Mode turns the same conversation into live voice. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/voice.md ### Connected work services The curated Composio integration includes Gmail, Calendar, Slack, Drive, Maps, and GitHub. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/composio.md ### Web search Enables best-effort web search and cautious retrieval of direct URLs. Source: https://github.com/Intelligent-Internet/genii ## Requirements and limits ### Development environment Requirement: Local setup requires Python 3.11+, Node.js 22+, and Docker for Postgres mode. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/setup.md ### External services and keys Requirement: The agent computer requires an E2B account with a desktop template and your own OpenAI API key. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/setup.md ### Not yet in production Limitation: The first milestone is intended for local development and does not yet promise production-ready self-hosting. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/roadmap.md ### Scope of this stage Limitation: Telegram, Discord, billing, and production use are excluded; iMessage is optional and beta. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/roadmap.md ### Backend must be running Limitation: v0.1 has no wake-up service when the local backend process is offline. Source: https://github.com/Intelligent-Internet/genii/blob/main/docs/community/setup.md ## Example workflows ### Try a calendar reminder - Your input: Set up a calendar trigger with test data in a local development environment. - Possible result: The trigger can prepare a message while the backend is actively running. - What to check: Consider the HTTPS callback, filters, and running backend service. ### Personal research briefing - Your input: Specify a topic and several approved URLs for a test run. - Possible result: The agent can retrieve content and create a briefing in chat. - What to check: Check sources, sandbox access, and runtime in the development milestone. ## Context The local development milestone requires E2B and OpenAI. Schedules require a running backend process; optional services have separate access and costs. README and community documentation reviewed again. Browser mode uses SQLite and requires E2B and OpenAI; Docker is part of Postgres mode. Voice and triggers require additional callback configuration. Production deployment remains outside this milestone. ## Sources - https://github.com/Intelligent-Internet/genii - https://github.com/Intelligent-Internet/genii/blob/main/docs/community/setup.md - https://github.com/Intelligent-Internet/genii/blob/main/docs/community/roadmap.md - https://github.com/Intelligent-Internet/genii/blob/main/docs/community/memory.md - https://github.com/Intelligent-Internet/genii/blob/main/docs/community/server-trigger-classification.md - https://github.com/Intelligent-Internet/genii/blob/main/docs/community/voice.md - https://github.com/Intelligent-Internet/genii/blob/main/docs/community/composio.md - https://e2b.dev/pricing - https://api.github.com/repos/Intelligent-Internet/genii [Sources and method](https://agenten-kompass.mmeierhoff.chatgpt.site/en/sources/)