Token Economics

​

Why Token Economics Matter

When you feed web data into an LLM, every token counts. Raw HTML is full of noise — navigation, ads, scripts, boilerplate — that wastes context window space, increases cost, and causes hallucination. WraithBytes delivers clean, structured outputs that strip away the noise. The result:

• Fewer tokens — Markdown and structured JSON use 60-90% fewer tokens than raw HTML
• Less hallucination — less noise in the context window means the model focuses on actual content
• Better results — structured data gives LLMs clear, unambiguous information to work with
• Lower cost — fewer input tokens means lower LLM API costs on top of already efficient acquisition pricing

​

Cost Efficiency

WraithBytes uses a simple token-based pricing model. Tokens are the universal unit of measurement for all web acquisition requests.

​

Token Costs

Standard acquisition includes fetching, parsing, and returning structured data (Markdown, JSON, links, images). AI summary generates a concise summary of the page content alongside the raw data. AI JSON extraction uses AI to extract structured data according to a custom schema. Image and audio transcription are currently in beta.

​

Free Tier

Every account includes 500 free tokens per month for testing and minimal use. Free tokens renew monthly and do not roll over.

​

Billing Principles

• Pay only for successful requests — failed fetches are never charged
• No subscriptions — no monthly commitments or pricing tiers
• No hidden fees — token costs are the only cost
• Linear scaling — your 1,000th request costs the same as your first
• Monthly renewal — free tier tokens reset each billing cycle

​

Cost Examples

​

What’s Next?