generated: '2026-07-19'
method: searched
source: https://docs.inceptionlabs.ai/get-started/models
billing_model: token-metered
currency: USD
unit: per 1M tokens
free_tier: 10,000,000 tokens on account creation
finops:
- model: mercury-2
  input: 0.25
  cached_input: 0.025
  output: 0.75
- model: mercury-edit-2
  input: 0.25
  cached_input: 0.025
  output: 0.75
notes: >-
  Cached input tokens are billed at a 90% discount ($0.025 vs $0.25 per 1M).
  Diffusion inference is positioned as roughly half the cost of comparable
  autoregressive frontier models. Enterprise pricing negotiated via sales.