Two inputs matter: official reference and current HaoAI price
Each model still follows input, output, cache, search, and media billing items. Pages show current HaoAI prices, official references, and calculated multipliers so teams can estimate real API cost quickly.
Live model-specific multiplier
The multiplier is calculated from current HaoAI prices and official reference prices. Input, output, and cache items are shown separately where available.
Pay for actual calls
No mandatory monthly plan and no package tiers. Token, cache, search, image, and other usage is billed from actual requests.
Traceable details
Every request can be reviewed with model, API key, token, latency, and cost details in the console.
Estimate bills with the current typical multiplier
These examples are generated from the typical multiplier currently available in the model catalog. Actual cost still depends on the model and request-level input, output, and capability usage.
HaoAI vs direct provider APIs
| Item | HaoAI | Direct provider API |
|---|---|---|
| Settlement multiplier | 0.15x | Official price × 1.00 |
| Integration | One API key for multiple model providers | Separate accounts and keys |
| Cost visibility | Request-level costs in one console | Split across provider consoles |
| Model switching | OpenAI / Anthropic compatible entry points | Separate provider integrations |
| Best fit | Multi-model usage, team cost control, fast experiments | Single-provider deep usage |
Check model prices, then create an API key
The model gallery lists billing items for each model. Once the cost is clear, create an API key in the console and start calling.