Definition
Token pricing charges based on the number of input and output tokens processed, common in AI and LLM products.
Why it matters
Token units are abstract for buyers, and costs can scale nonlinearly, so clarity is critical for margins and trust.
Pricing implications
Separate input and output rates, publish token-to-character guidance, and set minimums or tiers for heavy users.
Measurement tips
Track average tokens per request, model p90 usage, and monitor vendor pass-through costs.
Checklist
- Define what counts as a token and how it is measured.
- Publish example conversions for common use cases.
- Separate input vs output token pricing.
- Add usage tiers or volume discounts.
- Set rate limits to protect infrastructure.
- Monitor token mix and margin by cohort.
- Provide usage dashboards for transparency.
- Review pricing after model updates.
Examples
- .002 per 1K input tokens, .004 per 1K output tokens.