Summary of API usage fees for the latest OpenAI, VertexAI, and Claude, including GPT-4o mini
The GPT-4o mini appeared and surprised the world with its low API fee, but by the way, what is the difference in price and function from other models? That's what I'm worried about. So I looked up the latest API fees.
Comparison of OpenAI's GPT-4o mini and other models
OpenAI has priced GPT-4o mini as follows, making it 70-90% cheaper than GPT-3.5 Turbo and significantly more affordable than previous frontier models:
- Input: $0.15 / 1M Tokens
- Output: $0.60 / 1M tokens
Characteristics
GPT-4o mini has the following features, and it seems that it can be expected to enable the development and deployment of AI applications at a more affordable price as a high-performance and low-cost model.
- 128K Token Context Window
- Knowledge up to October 2023
- Scored 82% on the MMLU benchmark (outperformed GPT-4)
- Multimodal Support: Handles both text and vision inputs, with future plans to support text, image, video, and audio inputs and outputs.
- Accessibility: Available to ChatGPT Free, Plus, and Team users since July 18, 2024.
- API Integration: Accessible via the Assistant API, Chat Completion API, and Batch API for seamless integration into various applications.
API usage fee for each OpenAI model
Comparing the price of GPT-4o mini to other OpenAI models:
| Model | Input Cost ($/1M Tokens) | Output Cost ($/1M Tokens) |
| GPT-4o | $5.00 | $15.00 |
| GPT-4o mini | $0.15 | $0.60 |
| GPT-4 Turbo | $0.03 | $0.06 |
| GPT-4 | $0.03 | $0.06 |
| GPT-3.5 Turbo | $0.0015 | $0.002 |
Source：https://openai.com/api/pricing/
API usage fee for each model of Google Vertex AI
Even if you know the situation of OpenAI, you would like to compare the situation of other companies, so I summarized the situation of Generative AI on Vertex AI.
Basically, the fee is calculated per 1,000 characters, but 1,000 characters is not enough to compare, so I reorganized it in the table as a fee per million characters as a guide.
Please note:
If a query context is longer than 128K, all tokens are charged at long context rates.
| Model | Input fee (Per Million Characters) | Output fee (Per Million Characters) |
| Gemini 1.5 Pro | $0.3125 | $0.075 |
| Gemini 1.5 Flash | $0.01875 | $0.075 |
| Gemini 1.0 Pro | $0.3125 | $0.075 |
Source：https://cloud.google.com/vertex-ai/generative-ai/pricing
API usage fee for each model of Anthropic Claude
Finally, let's take a look at the API usage fee for each model of Anthropic Claude.
| Model | Input Cost ($/1M Tokens) | Output Cost ($/1M Tokens) |
| Claude 3.5 Sonnet | $3.00 | $15.00 |
| Claude 3 Opus | $5.00 | $20.00 |
| Claude 3 Sonnet | $2.50 | $12.50 |
| Claude 3 Haiku | $1.00 | $5.00 |
| Claude 2.1 | $4.00 | $16.00 |
| Claude 2.0 | $3.50 | $14.00 |
| Claude Instant 1.2 | $0.50 | $2.50 |
Source：https://www.anthropic.com/pricing#anthropic-api
Summary
The pricing differences among the models—Claude 3.5 Sonnet, GPT-4o mini , Gemini 1.5 Pro and others—highlight a wide range of options tailored to diverse needs. With varying costs, data update intervals, and performance capabilities, choosing the right model for your specific application is crucial. Switching between models based on your requirements ensures optimal performance and cost-effectiveness.
However, managing multiple platforms for registration, payments, and API key generation can quickly become overwhelming. That’s where RedPill steps in to simplify the process.
RedPill offers seamless access to over 200 top AI models, including OpenAI GPT-4o, Claude 3.5 Sonnet, Llama 2, and Mistral 7B, through a single, unified API. This integration eliminates the need for juggling multiple accounts, enabling you to access cutting-edge AI technology as new models are released.
Not only does RedPill maintain pricing parity with official platforms, but it also offers discounted credit packages, allowing you to enjoy these models at an even lower cost. With RedPill, you can effortlessly enhance your projects, save on expenses, and stay ahead in the rapidly evolving AI landscape—all while enjoying a streamlined and user-friendly experience.
