Skip to main content
A meter defines how you price and bill usage. Each meter has:

Meter slug

Unique identifier included in forward tokens to apply specific pricing

Fee model

Fixed cost per unit, percentage markup on provider cost, or both

Usage units

What you’re measuring — tokens, characters, seconds, or requests

Pricing tiers

Volume-based rate schedules that reward higher usage
Meters define pricing. Plans link to meters and provide the credit balance that usage draws from. See How Lava Monetize Works for the full picture of how meters, plans, and connections fit together.

Usage units

Usage units define what metric drives pricing. Set via the field:
Unit typeLabelRate basisExample use case
tokens_1mTokens$ per 1M tokensLLM APIs
characters_1mCharacters$ per 1M charactersText-to-speech
minutesSeconds$ per minuteAudio processing, voice calls
requestsRequests$ per requestImage generation, API calls

Fee models

The field determines how fees are calculated.
Charge a flat amount per usage unit, regardless of the provider’s cost:
fee = rate × units
Example: $2.00 per 1M tokens with input+output basis
  • 500 input tokens + 200 output tokens = 700 tokens
  • Fee: $2.00 × 700 / 1,000,000 = $0.0014

Pricing tiers

Tiers let you offer volume-based pricing. Each tier defines a usage threshold and rate. As cumulative usage increases, later tiers apply.
At least one tier is required. The first tier must start at 0, and tier starts must be in ascending order.

Example: Three-tier token pricing

TierStartRate (per 1M tokens)
10$5.00
21,000,000$3.00
310,000,000$1.00
For a connection that uses 5M tokens total:
  • First 1M at $5.00/1M = $5.00
  • Next 4M at $3.00/1M = $12.00
  • Total fee: $17.00
Tiers accumulate per connection. Each connection tracks total usage across all requests to determine the active tier.

Next steps