Bring Your Own API Key: You Control the AI
With RunTheAgents, you bring your own Anthropic or OpenAI API key. This gives you direct control over costs, model selection, and data policies.
What BYOK Means in Practice
Many AI platforms bundle the model cost into their pricing, marking it up and hiding the actual usage. You pay $50-200/month and have no idea how much of that goes to the AI model versus the platform's profit margin.
RunTheAgents uses a BYOK (Bring Your Own Key) model. You sign up with Anthropic or OpenAI directly, get your own API key, and enter it in your RunTheAgents dashboard. When your AI assistant makes a model call, it goes directly to your provider account. You see exactly what you are spending on AI usage through your provider's dashboard.
RunTheAgents charges a flat $25/month for managed hosting, infrastructure, monitoring, and updates. AI model costs are separate, transparent, and under your direct control.
Getting Your API Key
Step-by-step for both providers
Choose Your Provider
Decide between Anthropic (Claude models) and OpenAI (GPT models). Both are excellent. If unsure, see our guide on choosing an AI model. You can always switch later.
Create a Provider Account
Visit anthropic.com or platform.openai.com and create an account. Standard signup with email verification. Both providers offer free trial credits for new accounts.
Generate Your API Key
In your provider's dashboard, navigate to the API keys section and generate a new key. Give it a descriptive name like 'RunTheAgents Assistant.' Copy the key immediately; most providers only show it once.
Set a Spending Limit
Before using the key, set a monthly spending limit in your provider's billing settings. This caps your maximum AI model expense. Start conservative and increase as you understand your usage patterns.
Enter the Key in RunTheAgents
Paste your API key in the RunTheAgents dashboard. It is encrypted immediately and only used by your isolated instance. Your assistant is now connected to your chosen AI model.
Benefits of BYOK
Cost Transparency
You see exactly what you spend on AI model usage through your provider's dashboard. No markup, no hidden fees, no surprise charges. Your hosting fee and AI usage fee are completely separate and independently visible.
Model Control
You choose which model to use and can switch at any time. When new models are released, you can start using them as soon as your provider makes them available through their API.
Data Policy Control
Your data relationship is directly with your AI provider, not through an intermediary. You review and accept their terms directly. Any data handling questions go to your provider, not through a third party.
No Vendor Lock-In
Your API key works independently of RunTheAgents. If you ever want to self-host OpenClaw, your same API key works. If you want to switch hosting providers, your API key moves with you.
Understanding Your AI Costs
Light Usage (Personal Assistant)
A few conversations per day, occasional web research, and periodic task automation. Typical API cost: $5-10/month. Combined with $25 hosting: about $30-35/month total.
Moderate Usage (Business Assistant)
Regular client communication, daily research tasks, frequent browser automation. Typical API cost: $15-30/month. Combined with $25 hosting: about $40-55/month total.
Heavy Usage (Multi-Channel Operations)
High-volume messaging across multiple channels, frequent browser automation, extensive research. Typical API cost: $30-75/month. Combined with $25 hosting: about $55-100/month total.
Frequently Asked Questions
Related Pages
Ready to get started?
Deploy your own OpenClaw instance in under 60 seconds. No VPS, no Docker, no SSH. Just your personal AI assistant, ready to work.
Starting at $39.95/month. Everything included. 3-day money-back guarantee.