Our OpenAI API pricing calculator uses the information provided on OpenAI’s official website to estimate the total cost based on the number of word input prompts and the number of AI-generated words. According to OpenAI’s webpage, 1000 tokens generate about 750 words. We use that conversion rate for our calculator. (Pricing updated on February, 2025)
OpenAI API Pricing Calculator
$0.00
How OpenAI API pricing works
The OpenAI API provides a number of base language models that are priced based on tokens. These models consist of base language models (OpenAI o1, OpenAI o3-mini, GPT-4o, GPT-4o mini and the Legacy models (GPT-4 Turbo, GPT-4, GPT-3.5, and GPT base), fine-tuned language models (GPT-4o, GPT-4o mini, Legacy: GPT-3.5 Turbo, Babbage, Davinci, and Moderation), image models (DALL-E), and other models (Whisper and Embeddings). Each model has its distinct features and capabilities. OpenAI API uses token-based pricing, and every model has a different rate per 1000 tokens. We know it’s a bit confusing, but this article will give you a clear understanding of the costs involved in using the API.
OpenAI API pricing chart for each model
(Last updated: February 21, 2025)
| Language Models | Model | Input / 1k Tokens | Output / 1k Tokens |
|---|---|---|---|
| OpenAI o1 | $0.015 | $0.06 | |
| OpenAI o3-mini | $0.0011 | $0.0044 | |
| GPT-4o | $0.0025 | $0.01 | |
| GPT-4o mini | $0.00015 | $0.0006 | |
| Legacy Models | |||
| GPT-4 Turbo | GPT-4 Preview | $0.01 | $0.03 |
| GPT-4 Vision Preview | $0.01 | $0.03 | |
| GPT-4 | 8K context | $0.03 | $0.06 |
| 32k context | $0.06 | $0.12 | |
| GPT-3.5 Turbo 1106 | 16k context | $0.0010 | $0.0020 |
| GPT-3.5 Turbo Instruct | 4k context | $0.0015 | $0.0020 |
| Old Models* | Ada 001 | $0.0004 | |
| Babbage 001 | $0.0005 | ||
| Curie 001 | $0.002 | ||
| Davinci 003 | $0.020 | ||
| Fine-tuning models** | Babbage 002 | $0.0016 | $0.0016 |
| Davinci 002 | $0.0120 | $0.0120 | |
| GPT-3.5 Turbo | $0.0030 | $0.0060 | |
| Embedding models | Ada V2 | $0.0001 | |
| Audio models | Whisper | $0.006 | |
| Image models | DALL·E | 1024×1024 | $0.020 / image |
| 512×512 | $0.018 / image | ||
| 256×256 | $0.016 / image |
How to calculate OpenAI API cost using the number of words
Understanding the cost calculation for using the OpenAI API can be a bit tricky for some. The calculation is based on the number of tokens processed, where a token is approximately 4 characters or 0.75 words long. To determine the cost, you simply multiply the number of tokens by the price per 1,000 tokens. For instance, if you have used 10,000 tokens on GPT 3.5 Turbo 4k, the cost per 1,000 tokens is $0.002. Then, your total API usage cost would amount to $0.02.
According to OpenAI’s website, about 1,000 tokens are equal to roughly 750 words, or a ratio of 750 words per 1k tokens. Therefore, with our example scenario of using 10,000 tokens, you would be looking at 7,500 words. So, in this case, the price for utilizing the API would be $0.02 for around 7,500 words. Not bad, right?
But don’t worry about calculating your token usage and pricing; we’ve got you covered. Simply use our OpenAI API Pricing Calculator, which will handle all those calculations effortlessly for you.
Managing OpenAI API usage and cost
When you’re using the OpenAI API, it’s important to keep an eye on your usage and manage your expenses wisely. OpenAI offers tools that let you monitor your usage on a monthly basis using the Usage tab in your account. This feature presents your usage and costs in a format that makes it simple to see how your expenses are changing over time. By switching between cumulative views, you can gain insights into how you’re utilizing the API and take steps to stay within your budget.
How to view API usage
After login to your OpenAI API account, follow the steps below:
OpenAI API Price Changes
Updates
- Added – o1 and o1-mini September, 2024
- Added – GPT 4o mini July 18, 2024
- Added – GPT 4o (Omni) May 2024
- Added – GPT-4 Turbo Preview
- Added – GPT-4 Turbo Vision Preview
- Added – GPT-3.5 Turbo 1106 16K context (default) with $0.001/1k input and $0.002/1k output.
- Reduced – GPT-3.5 Turbo Fine-tuning reduced to $0.003/1k and $0.006/1k output.
The old pricing chart below expired on November 6, 2023
| Language Models | Model | Input / 1k Tokens | Output / 1k Tokens |
|---|---|---|---|
| GPT-4 | 8K context | $0.03 | $0.06 |
| 32k context | $0.06 | $0.12 | |
| GPT-3.5 Turbo | 4k context | $0.0015 | $0.002 |
| 16k context | $0.003 | $0.004 | |
| Old Models* | Ada 001 | $0.0004 | |
| Babbage 001 | $0.0005 | ||
| Curie 001 | $0.002 | ||
| Davinci 003 | $0.020 | ||
| Fine-tuning models** | Babbage | $0.0016 | $0.0016 |
| Davinci | $0.0120 | $0.0120 | |
| GPT-3.5 Turbo | $0.0120 | $0.0160 | |
| Embedding models | Ada V2 | $0.0001 | |
| Audio models | Whisper | $0.006 | |
| Image models | DALL·E | 1024×1024 | $0.020 / image |
| 512×512 | $0.018 / image | ||
| 256×256 | $0.016 / image |
FAQs
What is the difference between ChatGPT Plus and API?
ChatGPT Plus is a monthly subscription service offered by OpenAI for usage on chat.openai.com. The subscription allows you to use ChatGPT with faster speed plus a few more perks, such as access to DALL-E and other beta features. OpenAI’s API allows developers to integrate language models into their own applications or services. Updated: Users can use its browsing and analysis features. Our OpenAI API pricing calculator is updated accordingly.
What is the difference between GPT-4 API and ChatGPT Plus?
GPT-4 API allows you to access the GPT-4 language models (8k context or 32k context). ChatGPT Plus is a subscription to its chatbot with premium features. As of November 6, 2023, users can access the GPT-4 Turbo Preview with 128K context window!
Is the ChatGPT API a subscription?
No. API has its own pricing, which is a pay-as-you-use payment model.