GLM-4.5 API price
Z.AI’s published pricing lists GLM-4.5 at $0.60 per 1 million input tokens and $2.20 per 1 million output tokens, in USD. The same table lists cached input at $0.11 per 1 million tokens and cached-input storage as limited-time free. These rates are for the standard GLM-4.5 model—not GLM-4.5-Air. (docs.z.ai)
| GLM-4.5 charge | Z.AI listed price |
|---|---|
| Input tokens | $0.60 / 1M tokens |
| Cached input tokens | $0.11 / 1M tokens |
| Cached-input storage | Limited-time free |
| Output tokens | $2.20 / 1M tokens |
Z.AI’s pricing page says prices are in USD and presents these rates per million tokens. It does not give an end date for the limited-time storage offer in the captured pricing information, so don’t assume that offer will continue for a particular billing period. (docs.z.ai)
Estimate your GLM-4.5 API spend
For a basic estimate that excludes cached input, multiply input and output token volumes by their respective rates:
Estimated cost = (input tokens ÷ 1,000,000 × $0.60) + (output tokens ÷ 1,000,000 × $2.20).
For example, if an application uses 10 million input tokens and 2 million output tokens in a month, the estimate is:
- Input: 10 × $0.60 = $6.00
- Output: 2 × $2.20 = $4.40
- Estimated total: $10.40 for the month
This is a calculation from the published per-token rates, not a usage test or an invoice prediction. Actual budgeting depends on token usage and whether requests qualify for cached-input pricing. Z.AI lists cached-input and storage rates separately; apply those rates only when your usage and the provider’s billing rules make them applicable. (docs.z.ai)
GLM-4.5 is not GLM-4.5-Air
The names are easy to confuse, but the listed prices differ. Z.AI’s table lists GLM-4.5-Air at $0.20 per million input tokens and $1.10 per million output tokens, compared with GLM-4.5 at $0.60 and $2.20. Choose the row that matches the model ID you intend to call; do not use the Air rate to budget for GLM-4.5. (docs.z.ai)
Z.AI’s model documentation describes GLM-4.5 as a text-in, text-out model with a 128K-token context length and up to 96K output tokens. It also lists reasoning modes and tool invocation among its capabilities. These limits and features help identify the API model, but your expected bill is driven by the tokens actually used and the applicable pricing rates. (docs.z.ai)
Production API and pricing checks
Z.AI documents a chat-completions API using the model field and a bearer-token authorization header. Before forecasting spend, match the model ID in your requests to the intended model, review the provider’s current pricing table for any changes—especially the limited-time cached-input storage offer—and estimate input and output tokens from your expected workload. (docs.z.ai)
Sources
Z.AI Developer Documentation: Pricing