Pricing

诺玛AI uses pure pay-as-you-go pricing — you pay only for what you use. No subscriptions, no plans, no monthly fees, no minimum spend.

No subscriptions, no plans, no hidden fees. Top up and start using it right away, and your balance never expires.

How billing works#

Text models (billed per token)

Billing item

Description

Input Tokens

Tokens consumed by the input (prompt)

Output Tokens

Tokens consumed by the output (completion)

Cached Tokens

Cache-hit input tokens (discounted rate)

Other models

Type

Billing method

Image generation

Billed per image

Embeddings

Billed per input token

Getting started#

• Sign up — create an account in the 诺玛AI Console
• Top up — add credit as needed; your balance never expires
• Call — charged by actual usage, settled in real time

Viewing costs#

Console

• Sign in to the 诺玛AI Console
• Go to the Billing page
• View real-time cost breakdowns by model and by date

API response

Every API response includes a `usage` field so you can track usage easily:

{
  "usage": {
    "prompt_tokens": 100,
    "completion_tokens": 50,
    "total_tokens": 150
  }
}

Cost optimization tips#

• Choose the right model — use a lightweight model for simple tasks; there's no need to use a flagship model for every scenario
• Enable Prompt Caching — a repeated system prompt can cut input cost by 50–90%
• Control max_tokens — set a reasonable output length limit to avoid unnecessary token consumption
• Batch requests — combine multiple short requests into one to reduce repeated prompt overhead

For real-time per-model pricing, see the model catalog.

Last updated on June 23, 2026