Model Pricing
GeekAI supports hundreds of mainstream AI models through a unified interface. To implement our philosophy of low cost and high availability, we have categorized all AI models supported by our system based on official pricing, making it easier for developers to classify and select models:
| Model Tier | Price Range | Representative Models |
|---|---|---|
| Free Models | Free to use in all scenarios | glm-4-flash, huanyuan-lite, glm-4v-flash |
| B-Tier Models | API input price within 10 yuan/million tokens, Premium members can use for free on the platform | gpt-4o-mini, claude-3-5-haiku-latest, deepseek-chat |
| A-Tier Models | API input price between 10-40 yuan/million tokens, Professional members can use for free on the platform | gpt-4o, claude-3-7-sonnet-latest, deepseek-reasoner |
| S-Tier Models | Standard AI capabilities that can meet most needs, input price between 40-120 yuan | o1 |
| S+ Tier Models | AI models with prices exceeding 120 yuan | o1-pro, gpt-4.5-preview |
Disclaimer: The above classification is based solely on pricing dimensions and does not represent an evaluation of model quality, performance, or other dimensions.
At the same time, GeekAI’s AI model proxy service engine supports dynamically selecting the optimal API proxy for conversations at runtime to maximize reduction in API call costs. This optimization operates on two levels — one is price, and the other is availability. Low price without availability is not optimal, and availability with high price is also not optimal. The best solution means a dynamic balance between low price and availability.
Please visit the Model Marketplace page to view all models’ tier levels and minimum proxy prices. Different tier proxy prices (balanced, high availability, etc.) correspond to the proxy strategy selected when creating an API KEY, and automatically switch to higher-tier strategies when lower-price strategies are unavailable to ensure availability.
API interface calls currently only support payment with platform credits. If you receive an insufficient balance prompt when calling the API, please recharge your credits through the Purchase Plans page (you can check your remaining balance on the Available Credits page) or invite friends to register to get free credits.
For API call rate limits, please refer to the rate limit field in the Model Marketplace. A field value of ”-” indicates no rate limit. Exceeding the rate limit will return a 429 error code, so please control your call frequency accordingly.