Frequently Asked Questions

Mercury deploys a private, always-on AI agent on dedicated infrastructure you control. We handle the full launch stack — server provisioning, software installation, channel configuration, and production hardening — so you get a fully working agent without touching a command line.

Unlike ChatGPT or other hosted AI services, Mercury runs on your own dedicated server. You choose how to power it: use Mercury Credits for instant access with no API accounts needed, or bring your own API keys (BYOK) for direct provider control. Either way, you get full data ownership, no usage caps imposed by us, and access to any AI model — Claude, GPT, Gemini, Kimi K2.5, MiniMax, and even local models via Ollama. Mercury is always-on, integrates with Telegram/Signal/WhatsApp, and proactively works for you 24/7.

Not at all. Most people never want to deal with cloud servers, Docker, channel setup, API keys, uptime, or security hardening. That's exactly what Mercury handles for you. You tell us how you want your agent configured, and we deliver a fully working setup without the DevOps overhead.

Mercury supports all major AI model providers: Claude (Anthropic), GPT (OpenAI), Kimi K2.5 (Moonshot), MiniMax M2.5, Gemini (Google), and Ollama for running local/free models. Access them instantly via Mercury Credits (no API accounts required), or connect your own API keys if you prefer direct control. Switch between models freely — you're never locked into a single provider.

Agentic engineering is the practice of directing AI systems to plan, execute, and improve real work on your behalf. Think of it as turning an AI model into an operational teammate instead of a simple chatbot. With Mercury, you do not need to master that setup yourself — we deliver the infrastructure and configuration so your assistant is ready from day one.

Head to our checkout page, pick your plan, and your cloud deployment typically finishes in about 10–15 minutes after payment.

Cloud deployment is fully automated and usually takes about 10–15 minutes. When you complete checkout, our system automatically provisions a dedicated cloud server, installs Docker and Mercury, configures your messaging channels, and delivers your login credentials — all without any human involvement.

We offer three deployment options to fit your needs: Cloud — We deploy Mercury on a dedicated cloud server that we provision and maintain for you. Fully automated, usually ready in about 10–15 minutes. Best for most users who want a hassle-free AI agent setup. On-Site ($1,500) — We physically come to your location and set up Mercury on your own hardware or network. Scheduled based on availability. Ideal for businesses with strict data residency requirements. Ship-In ($1,000) — You ship us your hardware (mini PC, NUC, old laptop, etc.) and we install, configure, and ship it back ready to plug in. Turnaround: 5–10 business days.

Mercury supports Telegram, Discord, WhatsApp, Signal, iMessage, Slack, and Google Chat. You can connect one or multiple channels simultaneously, so your personal AI assistant is reachable wherever you prefer to chat.

Absolutely. If you start with a Cloud deployment and later want to move to your own hardware (or vice versa), we can migrate your setup. Contact us and we'll work out a migration plan that preserves your configuration and data.

For Cloud plans, everything is automated. After checkout, our system provisions your dedicated cloud server, installs and configures Mercury with Docker, connects your messaging channels, and emails you your login credentials and setup instructions. Most cloud deployments finish in about 10–15 minutes. For On-Site deployments, we'll schedule a visit based on availability. For Ship-In, you ship us your hardware and we configure and return it within 5–10 business days.

Cloud deployments run on high-performance dedicated cloud VPS infrastructure. The exact specs depend on your plan tier, but all plans include dedicated CPU cores, SSD storage, and generous RAM to ensure your AI agent runs smoothly 24/7. Our infrastructure is selected for reliability, low latency, and strong uptime guarantees.

Our Cloud hosting plans are: Starter $29/month, Pro $49/month, and Business $89/month. For On-Site deployment, it's $1,500 full payment upfront (customer pays travel expenses). Ship-In deployment is $1,000 full payment upfront (customer pays all shipping costs). Visit our Pricing page for full details.

Every Cloud plan includes full Mercury installation, automated server provisioning and hosting, your choice of messaging channel integration, ongoing server maintenance, 30-day post-launch support, and access to our support team. Higher tiers include additional features like priority support, more server resources, multiple channel connections, and advanced configuration options. Check our Pricing page for a detailed comparison.

Cloud plans: Cloud plans can be cancelled anytime — you keep access until the end of your billing period.

We accept all major credit and debit cards (Visa, Mastercard, American Express), as well as other payment methods available through our secure Stripe-powered checkout. For On-Site and Ship-In deployments, we can also arrange invoicing for business customers.

No hidden fees — what you see on our pricing page is what you pay. You have two options for AI model access: Mercury Credits (buy credits from us, no API accounts needed — easy one-bill experience) or BYOK (connect your own API keys and pay providers directly at their standard rates). Credits include a small convenience markup; BYOK is pure pass-through. Both options give you full transparency over your AI spending.

Yes. Your Mercury instance runs on a dedicated server — your data is not shared with other users. All connections are encrypted, and we follow security best practices for server hardening, access control, and software updates. For Cloud deployments, your server is hosted on enterprise-grade cloud infrastructure with strong physical and network security.

No. If you use BYOK, your API keys are stored only on your own server and used directly by Mercury to communicate with AI providers. We never store, log, or access your keys after setup. If you use Mercury Credits, no API keys are involved on your end at all — we handle provider access on our infrastructure. Either way, you maintain full control.

For Cloud deployments, your data is hosted on a dedicated cloud VPS provisioned specifically for you. The server location depends on availability and your preference — we can discuss region options during setup. For On-Site and Ship-In deployments, your data stays entirely on your own hardware at your own location.

Only you (and anyone you explicitly authorize) has access to your Mercury AI agent. During initial setup, our automated system configures everything, but once deployed, you control all access. You can set up user permissions, restrict which messaging accounts can interact with your agent, and manage everything through the Mercury dashboard.

Cloud deployments run on enterprise-grade infrastructure with 99.9% uptime target. Business-tier customers receive a formal SLA with response-time guarantees. We monitor all deployments 24/7 and proactively address issues before they affect your agent. Enterprise customers can negotiate custom SLA terms including dedicated support response windows and escalation paths. Contact our enterprise team for SLA details.

Yes. Every Mercury deployment includes comprehensive audit logging: agent actions, tool calls, API requests, message routing, and configuration changes are all logged with timestamps. Business and Enterprise customers can access detailed audit logs through the dashboard or export them for compliance review. Enterprise deployments can integrate with your existing SIEM or logging infrastructure.

BYOK means you connect your own AI provider API keys (OpenAI, Anthropic, Google, Moonshot, etc.) and pay providers directly at their standard rates. Your keys are stored only on your own server — we never see, store, or log them after initial setup. You can rotate keys anytime through your Mercury dashboard. BYOK gives you full cost transparency and provider independence. Alternatively, Mercury Credits lets you access all providers through a single billing relationship with us — no API accounts needed.

For Cloud deployments, we can provision your server in specific regions based on availability. On-Site deployment ($1,500) puts everything on your own hardware at your location — your data never leaves your premises. Ship-In deployment ($1,000) also results in fully local operation. For regulated industries requiring data residency guarantees, contact our enterprise team to discuss regional hosting options and contractual data residency commitments.

Mercury is architected for compliance-friendly deployment: dedicated servers (no multi-tenant data mixing), encrypted connections, BYOK key isolation, and comprehensive audit logging. On-Site and Ship-In deployments give you complete infrastructure control, which simplifies compliance posture. For customers requiring formal SOC 2 or HIPAA BAA agreements, our enterprise team can discuss compliance documentation and contractual guarantees.

We maintain a documented incident response process. All deployments are monitored for anomalies. In the event of a detected issue: (1) affected customers are notified immediately, (2) containment measures are applied, (3) a root-cause analysis is conducted, (4) a remediation report is provided. Enterprise customers receive priority incident response with guaranteed notification windows defined in their SLA.

You own your data completely. On Cloud deployments, your conversation history and agent data are stored on your dedicated server. You can delete data at any time through your Mercury dashboard or directly on your server. If you cancel your plan, you retain access until the end of your billing period, after which your server is decommissioned. On-Site and Ship-In customers have full physical control — data retention is entirely your decision.

Yes. Enterprise deployments support parent organization accounts with centralized billing, seat issuance, staged rollout, and governed multi-level AI deployments. A parent account can manage multiple agent deployments across teams, control BYOK policies at the org level, and enforce usage guardrails. Contact us for enterprise pricing and deployment options.

Enterprise timelines depend on scope. Standard Business-tier Cloud deployments are live in 10–15 minutes. Enterprise multi-seat deployments typically take 1–2 weeks for initial rollout (including org setup, BYOK configuration, and staged seat provisioning). On-Site enterprise deployments are scheduled based on availability and travel logistics. We provide a deployment timeline and checklist during the enterprise scoping call.

Enterprise customers receive: priority support with guaranteed response windows, dedicated onboarding for multi-seat rollouts, custom SLA with uptime and incident-response terms, quarterly reviews covering usage, optimization, and expansion opportunities, and direct access to our engineering team for integration and configuration questions. Request enterprise support details.

All plans include 30 days of post-launch support. You can reach us through multiple channels: email us at support@mercuryinstall.com, join our Discord community for real-time help, or use our Contact page to submit a detailed inquiry. We typically respond within 24 hours.

We offer dedicated support hour packages for customers who need hands-on assistance beyond the included 30-day post-launch support. Support hours can be used for advanced configuration, troubleshooting, custom integrations, training, and more. Higher plan tiers come with discounted support hour rates. Visit our Support page for details on packages and pricing.

Absolutely! You can email us anytime at support@mercuryinstall.com or reach out through our social channels including Telegram, X (Twitter), and Discord. For urgent matters, Discord is often the fastest way to get a response.

Yes. For businesses and organizations with advanced requirements — such as multi-user deployments, custom integrations, SLA guarantees, or dedicated support — we offer enterprise-grade AI automation solutions. Contact us to discuss your needs and we'll put together a tailored plan.