OpenClaw Integration

OpenClaw is a self-hosted gateway that connects your favorite chat apps—WhatsApp, Telegram, Discord, iMessage, and more—to AI coding agents. With Halfbill as a custom provider, you can route all your OpenClaw traffic through a single API, use Claude and GPT models, and keep full visibility into usage and costs.

Setup

export HALFBILL_API_KEY=llmgtwy_your_api_key_here

{
	"models": {
		"mode": "merge",
		"providers": {
			"halfbill": {
				"baseUrl": "https://api.halfbill.uk/v1",
				"apiKey": "${HALFBILL_API_KEY}",
				"api": "openai-completions",
				"models": [
					{
						"id": "gpt-5.5",
						"name": "GPT-5.5",
						"contextWindow": 400000,
						"maxTokens": 128000
					},
					{
						"id": "claude-opus-5",
						"name": "Claude Opus 5",
						"contextWindow": 200000,
						"maxTokens": 32000
					},
					{
						"id": "claude-haiku-4-5",
						"name": "Claude Haiku 4.5",
						"contextWindow": 200000,
						"maxTokens": 32000
					}
				]
			}
		}
	},
	"agents": {
		"defaults": {
			"model": {
				"primary": "halfbill/gpt-5.5"
			}
		}
	}
}

Why Use Halfbill with OpenClaw

• Model flexibility — Switch between GPT-5.5, Claude Opus 5, Claude Haiku 4.5, or any other Halfbill model
• Cost tracking — Monitor exactly how much your chat agents cost to run
• Single bill — No need to manage multiple API provider accounts
• Response caching — Repeated queries hit cache, reducing costs

Switching Models

Change the primary model in your config to switch between any model:

{
	"agents": {
		"defaults": {
			"model": { "primary": "halfbill/claude-opus-5" }
		}
	}
}

Model Fallback Chain

OpenClaw supports fallback models. If the primary model is unavailable, it automatically falls back:

{
	"agents": {
		"defaults": {
			"model": {
				"primary": "halfbill/gpt-5.5",
				"fallbacks": ["halfbill/claude-opus-5"]
			}
		}
	}
}

Available Models

Halfbill uses canonical model IDs. You can use any model from the models page. Flagship models include:

Tips for Chat Agents

Optimize Costs

1. Use smaller models for simple tasks — Claude Haiku 4.5 handles basic Q&A well
2. Enable caching — Halfbill caches identical requests automatically
3. Set token limits — Configure max tokens to prevent runaway costs

Improve Response Quality

1. Choose the right model — Claude Opus excels at nuanced conversation, GPT-5.5 at general tasks
2. Use system prompts — Configure your agent's personality and capabilities
3. Test multiple models — Halfbill makes it easy to A/B test Claude against GPT