ClawServe API vs BYOK: Which AI Connection Mode is Right for You?

Two Ways to Power Your AI

When you deploy an OpenClaw instance on ClawServe, you choose how it connects to AI models. We offer two modes:

ClawServe API Mode

In this mode, your instance routes AI requests through our managed proxy. We handle the API keys, rate limits, and provider failover.

Pros:

• No API keys to manage
• Access to multiple models (Claude, GPT-4, DeepSeek, Kimi) from one interface
• Pay-as-you-go pricing with transparent token costs
• Automatic failover if a provider has downtime
• Usage tracking and spending limits built in

Cons:

• Slightly higher per-token cost (we add a small margin for infrastructure)
• Requests route through our proxy (adds ~50ms latency)

Best for: Users who want simplicity and don't want to manage API keys.

BYOK (Bring Your Own Key) Mode

In this mode, your instance connects directly to the AI provider using your own API keys.

Pros:

• Direct connection to the provider (lowest latency)
• Use your existing API credits and volume discounts
• No markup on token pricing
• Full control over which provider and model to use

Cons:

• You manage your own API keys and rate limits
• No automatic failover between providers
• You handle billing directly with each provider

Best for: Developers and power users who already have API keys and want maximum control.

Switching Between Modes

You can switch between ClawServe API and BYOK at any time from the Models section of your Dashboard. Your conversation history and skills are preserved — only the AI backend changes.

Our Recommendation

• New to AI? Start with ClawServe API — it's the easiest way to get started.
• Have API keys? BYOK gives you the best price and lowest latency.
• Want both? You can switch anytime depending on the task.