Frequently Asked Questions

What AI providers does noHuman Team support?

noHuman Team works with 10+ AI providers out of the box, giving you full flexibility over which models power your noHumans. Supported providers include Anthropic Claude (Claude Opus, Sonnet, Haiku), OpenAI GPT (GPT-4o, o1, o3), Google Gemini (Gemini 2.5 Pro, Flash), DeepSeek, Mistral, Groq (for ultra-fast inference), and xAI Grok. You can also connect through OpenRouter, which unlocks access to 100+ additional models from dozens of providers — all through a single API key.

Each noHuman on your team can run a different model, so you can assign expensive reasoning models to complex tasks and fast, affordable models to routine work. This lets you optimize both quality and cost across your noHuman team. Just add your API keys in the setup wizard and choose which model each noHuman uses. → See Setup Wizard docs

Where does my data go? Is noHuman Team private?

Your data stays on your machine. noHuman Team runs entirely on your local computer using Docker containers — there is no noHuman Team cloud server receiving your conversations, files, or API keys. When your noHumans call an AI provider (like Anthropic or OpenAI), those API requests go directly from your machine to the provider using your own API keys. We never proxy, store, or have access to those calls.

Your workspace files, noHuman memory, and team conversations all live in a local folder on your hard drive. This architecture means noHuman Team offers maximum data privacy for noHuman workflows — you get the power of a full noHuman team without sending sensitive business data to a third-party platform. If privacy and data sovereignty matter to your workflow, local-first noHumans are the way to go. → Learn about Environment Variables & API Keys

Can I use noHuman Team without an API key?

You need at least one AI provider API key to use noHuman Team. The noHumans are powered by large language models from providers like Anthropic, OpenAI, or Google, and those providers require API keys for access. The good news: you can start with a single API key from any supported provider (even a free-tier key from Google Gemini or Groq) and assign it to all your noHumans.

As your usage grows, you can add keys from multiple providers to distribute costs or access specialized models. noHuman Team itself has no usage fees beyond the one-time purchase — you only pay for the AI tokens you actually consume through your chosen providers. This gives you full control over your AI spending with no surprise subscription charges. → Getting Started guide

How much does noHuman Team cost?

noHuman Team is a one-time purchase of $149 — no subscriptions, no per-seat fees, no token markups. That includes the full desktop application, all team templates, and 6 months of free updates. You pay once and own it. Compare that to SaaS noHuman platforms that charge $50–200/month per user.

After your update period, the app keeps working indefinitely — you just won't receive new features or patches unless you renew. The only ongoing cost is your AI provider API usage (Anthropic, OpenAI, etc.), which you pay directly to those providers at their standard rates. For most small teams and solopreneurs, that's $5–50/month in API costs depending on usage volume and model selection. → See pricing details

What happens after the 6-month update period?

After your 6 months of included updates expire, noHuman Team keeps working exactly as it did. You don't lose access, your noHumans don't stop, and nothing gets locked behind a paywall. You simply stop receiving new feature releases and patches. If you want to continue getting updates after that, you can purchase an update renewal at a discounted rate.

Many users find that the version they have works perfectly and don't need to renew immediately. This model means you're never locked into recurring payments just to keep your noHuman team running — a major advantage over subscription-based AI platforms.

What is the refund policy?

noHuman Team offers a 14-day money-back guarantee. If you purchase the app and decide it's not right for you within 14 days, contact support for a full refund — no questions asked. We want you to have enough time to set up your team, test different configurations, and see real results before committing.

Most users know within the first few hours whether noHuman Team fits their workflow, but you have two full weeks to be sure. After the 14-day window, all sales are final since the product is a downloadable desktop application.

What team templates are available?

noHuman Team ships with pre-built team templates designed for common workflows. The Startup template gives you a CEO noHuman that coordinates a Developer, Marketer, and Automator — ideal for solopreneurs who need a full founding team. The Dev Squad template sets up a Tech Lead with Architect, Coder, and QA noHumans for software engineering teams. The Content Factory template creates a Content Director with Writer, Editor, and SEO noHumans for content production at scale.

Each template comes with pre-configured roles, communication rules, and workflow patterns that you can use as-is or customize. Templates are the fastest way to go from download to a working noHuman team in under 5 minutes. → See noHuman Management docs

Can I customize noHuman roles?

Absolutely. Every noHuman role in noHuman Team is fully customizable. You can start from a preset role (Developer, Marketer, Designer, Researcher, etc.) and modify its system prompt, responsibilities, communication style, and tool access. Or you can create completely custom roles from scratch — define what the noHuman does, how it communicates, and what tools it can use.

Want a Legal Analyst noHuman? A Customer Support bot? A Data Scientist? Just describe the role and the noHuman will operate accordingly. You can also change roles after your team is running — swap a Marketer for a Sales noHuman, or adjust the CEO's delegation style. This flexibility means noHuman Team adapts to any business workflow, not just the templates it ships with. → Customize in the Dashboard

How does Telegram integration work?

noHuman Team connects to Telegram for mobile communication with your noHumans. After setup, you can DM your team leader directly in Telegram — send tasks, ask questions, or check status from your phone. Your team leader (typically the CEO noHuman) receives your message and delegates to the right team member automatically.

You can also create a Telegram group chat where all noHumans participate, letting you observe team discussions in real time. The integration supports voice messages too — speak your instructions and the noHuman transcribes and processes them. This means you can manage your noHuman team while commuting, in meetings, or away from your desk. Telegram becomes your mobile command center for your entire AI workforce. → Team Communication docs

What is coding delegation?

Coding delegation is one of noHuman Team's most powerful features. When a noHuman needs to write or edit code, it can spawn a dedicated coding sub-agent — like Claude Code, Codex, or other AI coding assistants — that runs inside the noHuman's own sandboxed container.

This means your Developer noHuman doesn't just write code in its chat context (which burns tokens fast). Instead, it delegates to a specialized coding tool that can explore files, run tests, iterate on errors, and deliver working code. The coding sub-agent operates with full filesystem access inside the sandbox, can install packages, run build tools, and verify its own output. Once done, it reports back to the parent noHuman. This architecture gives you professional-grade AI coding without the context window limitations of a single LLM conversation. → Advanced Configuration docs

How do noHumans communicate with each other?

noHumans in noHuman Team communicate through an internal message bridge — a lightweight routing system that runs alongside your noHuman containers. Each noHuman has a unique ID and can send direct messages to any other noHuman or broadcast to the whole team. The bridge supports auto-routing: a noHuman can send a message without specifying a recipient, and the bridge intelligently delivers it based on role and context.

For example, if the Marketer noHuman writes landing page copy and sends it to the bridge, it automatically routes to the Developer for implementation. All inter-noHuman communication is logged, so you can review the full conversation history. noHumans follow team manifests — shared rules that define communication protocols, handoff procedures, and quality standards. This structured communication is what makes multi-agent teams actually productive instead of chaotic. → Team Communication docs

How does noHuman memory and persistence work?

Each noHuman in noHuman Team maintains persistent memory across sessions through a file-based system. Every noHuman has a STATUS.md file that stores its current task, completed work, blockers, and next steps — this is the first thing it reads on startup. noHumans also write daily notes (memory/YYYY-MM-DD.md) that log what happened during each session, and a curated long-term memory file (MEMORY.md) that stores important decisions, preferences, and lessons learned.

When a noHuman's context window fills up, a smart compaction process summarizes the conversation and preserves critical information, so the noHuman doesn't lose track of what it was doing. This means your noHuman team members remember context between sessions — they know what they worked on yesterday, what decisions were made last week, and what your preferences are. It's the closest thing to giving noHumans a real working memory.

Can I assign different AI models to different noHumans?

Yes — per-noHuman model selection is a core feature of noHuman Team. You can assign any supported AI model to each noHuman independently. This enables powerful cost optimization strategies: use a high-capability model like Claude Opus or GPT-4o for your CEO noHuman that needs complex reasoning, a fast mid-tier model like Claude Sonnet or Gemini Flash for your Developer, and an affordable model like DeepSeek or Groq for routine tasks.

You can change models at any time without reconfiguring the noHuman's role or memory. This granular control means you can run a full noHuman team for a fraction of what it would cost to run every noHuman on the most expensive model. Most users find their sweet spot at $10–30/month in API costs by mixing models strategically. → Environment Variables & API Keys

What is the heartbeat/proactive system?

The heartbeat system lets your noHumans work proactively without being asked. At configurable intervals, each noHuman receives a "heartbeat" pulse that triggers it to check for pending work. noHumans can be configured to check email inboxes for urgent messages, scan calendars for upcoming events, monitor social media mentions, check weather for relevant updates, or perform any recurring task you define.

The noHuman decides whether something needs attention — if yes, it notifies you; if not, it stays quiet. You control what each noHuman checks and how often through a simple HEARTBEAT.md configuration file. This transforms your noHuman team from reactive assistants into proactive colleagues who surface important information before you have to ask. It's the difference between an AI that answers questions and an AI that anticipates needs. → Advanced Configuration

How does smart context management work?

Large language models have limited context windows, and multi-agent conversations can fill them fast. noHuman Team uses smart compaction to manage this automatically. When a noHuman's conversation history approaches its context limit, the system summarizes older messages into a condensed format, preserving key decisions, task states, and important details while discarding routine back-and-forth.

This means your noHumans can work on long-running tasks for hours without losing track of what they're doing or running out of context space. Combined with the file-based memory system (STATUS.md, daily notes, MEMORY.md), noHumans maintain continuity even through multiple compaction cycles. The result: you save 40–60% on API token costs compared to naive approaches that resend entire conversation histories. Smart context management is what makes running a multi-agent team economically viable.

Can I access noHuman virtual desktops?

Yes. Each noHuman in noHuman Team runs in its own Docker container with a full Linux desktop environment. You can connect to any noHuman's desktop via VNC (Virtual Network Computing) directly from the noHuman Team dashboard — no additional software needed. This lets you watch noHumans work in real time, see their browser sessions, inspect files, or even interact with their desktop manually.

Virtual desktops are especially useful for debugging noHuman behavior, verifying that a coding noHuman's build succeeded, or checking what a research noHuman found in its browser. Each container is fully sandboxed, so noHumans can install packages, run applications, and browse the web without affecting your host system or other noHumans. It's like having remote desktop access to each team member's workstation. → Virtual Desktops docs

How does the shared workspace work?

All noHumans on your noHuman team share a single workspace folder that's mounted into every container. When the Marketer writes a content brief, the Developer can immediately see it. When the Developer pushes code, the CEO can review it. This shared folder is synced to a directory on your host machine, so you can also access, edit, or add files from your own computer using any editor or file manager.

Despite the shared access, each noHuman's container is sandboxed — noHumans can't access each other's system files, install malware on your host, or interfere with other noHumans. The workspace is the collaboration layer; the container is the security boundary. This design mimics how real teams work: shared Google Drive for collaboration, but separate laptops for security. → Getting Started docs

What's the difference between noHuman Team and OpenClaw?

OpenClaw is the open-source AI agent runtime that powers each individual agent — it handles model connections, tool execution, memory, and communication for a single agent. noHuman Team is a desktop application built on top of OpenClaw that orchestrates multiple noHumans into a coordinated team.

Think of it this way: OpenClaw is the engine; noHuman Team is the car. With raw OpenClaw, you'd need to manually configure Docker containers, set up networking between noHumans, manage the message bridge, and write team manifests. noHuman Team wraps all of that into a point-and-click desktop app with a setup wizard, dashboard, team templates, and visual management tools. If you're technical and want maximum control, OpenClaw is free and open source. If you want a working noHuman team in minutes without DevOps, noHuman Team is the product. → OpenClaw on GitHub

What are the system requirements?

noHuman Team requires Docker Desktop (available for Windows, macOS, and Linux) and a stable internet connection for AI provider API calls. Recommended specs: 12GB RAM minimum (24GB recommended for teams of 4+ noHumans), any modern multi-core CPU, and 50GB of free disk space for container images. Each noHuman container uses approximately 3–6GB of RAM when active.

You don't need a GPU — all AI inference happens in the cloud via your API keys. noHuman Team runs on Windows 10/11, macOS 12+, and Ubuntu/Debian Linux. Installation takes under 5 minutes: install Docker Desktop, run the install script, and follow the setup wizard. → Getting Started guide

Can I use noHuman Team offline?

noHuman Team requires an internet connection for AI model API calls — since your noHumans use cloud-based LLMs (Anthropic, OpenAI, Google, etc.), they need network access to think. However, the application itself, the Docker containers, the message bridge, and the noHuman orchestration all run locally on your machine.

If you configure a local LLM (like Ollama or LM Studio) as your provider via an OpenAI-compatible endpoint, you could theoretically run fully offline — though local model quality is significantly lower than cloud models for complex noHuman tasks. In practice, most users need internet access, but your data, files, and configurations remain local regardless.

How do updates work?

noHuman Team updates are delivered through the desktop application. When a new version is available, you'll see a notification in the dashboard and can update with one click. Updates include new features, bug fixes, improved team templates, and compatibility updates for new AI models and providers. Your 6-month update period starts from the date of purchase.

During that period, all updates are free. After it expires, your current version continues to work indefinitely — you just won't receive new releases. You can purchase an update renewal at any time to get the latest version and restart the update clock. Your team configurations, noHuman memories, and workspace files are preserved across updates — nothing gets lost when you upgrade.

What are common business use cases for noHuman Team?

noHuman Team is built for solopreneurs, small teams, and businesses that want AI-powered productivity without hiring. Common use cases include: Startup operations — a CEO noHuman coordinates product development, marketing, and automation across your noHuman team. Content production — an editor delegates writing, SEO optimization, and social scheduling to specialized noHumans. Software development — a lead dev noHuman manages frontend, backend, and QA noHumans with coding delegation for real code output.

Research and analysis — noHumans browse the web, synthesize reports, and deliver briefings on competitors, markets, or trends. Customer support automation — noHumans draft responses, categorize inquiries, and escalate complex cases. Personal productivity — a single noHuman team handles email triage, calendar management, and task coordination. The key advantage: each noHuman runs in its own sandboxed environment with its own browser and tools, so they can actually do work — not just suggest what you should do. → Dashboard Overview