Pay per token, not per hour
No subscriptions. No idle costs. You only pay for the AI tokens your agents actually use.
Model Pricing
Price per 1M tokens · 1.5x markup on provider rates for infrastructure costs
| Model | Provider | Input / 1M tokens | Output / 1M tokens |
|---|---|---|---|
| Claude Sonnet 4 | Anthropic | $4.50 | $22.50 |
| Claude Haiku 4.5 | Anthropic | $1.20 | $6.00 |
| GPT-4o | OpenAI | $3.75 | $15.00 |
| GPT-4o Mini | OpenAI | $0.22 | $0.90 |
| Gemini 2.0 Flash | $0.15 | $0.60 |
Prices shown include the 1.5x infrastructure markup. Provider base rates are lower.
Credit Packages
Buy more, get bonus credits. Custom amounts also accepted ($20+).
Frequently asked questions
- How does token-based billing work?
- Every AI model call through Clawfy is metered by tokens (the units of text processed). You pay per token used — input tokens (your prompt) and output tokens (the response). Credits are deducted in real-time as you use them.
- How do credits work?
- Buy credits upfront, then credits are deducted per API call based on tokens used. 1 credit ≈ $0.01. You only pay for what you actually use.
- Do credits expire?
- No, credits never expire. Use them whenever you want.
- What happens when I run out of credits?
- API requests will be rejected and running agents will be stopped automatically. Top up credits to resume.
- Why is there a markup on model prices?
- The 1.5x markup covers infrastructure costs (proxy hosting, monitoring, scaling) and gives you a unified API for all providers with a single API key.
- What AI models do you support?
- We support Claude (Anthropic), GPT (OpenAI), and Gemini (Google) models through a single API endpoint. More models are added regularly.
- Can I use my own API keys instead?
- Clawfy provides a unified proxy with a single key for all providers. This simplifies management and lets you switch models without changing keys.