generated: '2026-09-10'
method: searched
source: https://modelrush.ai/pricing
plan_count: 0
model: pay-as-you-go
note: ModelRush publishes no subscription tiers -- pricing is pure usage-based,
  per model, per operation, per region, denominated in USD, generated from the
  same versioned registry as the gateway. Prepaid credit is purchased via Stripe
  Checkout in $10, $50, and $100 packs. "Evaluation requests record real usage
  without deducting credits."
billing:
  currency: USD
  credit_packs: [10, 50, 100]
  units:
  - surface: chat/text
    unit: per 1M tokens (input and output priced separately)
  - surface: text-to-speech
    unit: per character
  - surface: audio transcription
    unit: per second
  - surface: image generation/editing
    unit: per generation
  - surface: video generation
    unit: per second (resolution-dependent)
  - surface: voice management
    unit: per operation
regional_variation: true
examples:
- model: deepseek-v4-flash
  region: cn-beijing
  input_per_1m_tokens_usd: 0.4664
- model: deepseek-v4-flash
  region: sg-singapore
  input_per_1m_tokens_usd: 0.484
- model: qwen-3.7-max
  region: cn-beijing
  input_per_1m_tokens_usd: 1.815
- model: qwen-3.7-max
  region: sg-singapore
  input_per_1m_tokens_usd: 2.75
machine_readable:
  note: GET https://api.modelrush.ai/v1/models answers keyless with the live model
    catalog and registry revision; the pricing page is generated from the same
    registry.