Team cost & model comparison
MaaS model comparison
The same workload priced on every selectable MaaS model, each with its real period split.
Competitor comparison
Grouped by class — flagship class, then the pricier flash class. The MaaS anchor row is highlighted red.
12-month cost projection
Each line prices your exact workload on that model's list rates, with your growth rate applied.
Cost of a single API request
Price one request by its exact input, cache-hit, and output tokens.
Model
Request
Billing period
Estimated cost · GLM-5.3$0.00
Reference pricing & billing formula USD per 1 million tokens
Documentation list prices are in USD; when BRL is selected, values are converted at your editable rate.
Cost =
(input − cached) × input price + cached × cache-hit price + output × output price, each per 1M tokens. Only successfully generated content is billed. Integer scaling (×1000) avoids floating-point drift. Huawei bills by Beijing time: Period 1 = 08:00–20:59, Period 2 = 21:00–07:59 (30% cheaper).
DeepSeek-V4-Flash — 3M input / 7M output (flat rate)$2.295
GLM-5.2 · Period 1 — 1.5M input incl. 0.5M cached / 3M output$14.73
Scenario assumptions
Export the full page
Capture the entire calculator — controls, KPIs, comparisons and projections — as one PNG.