How much does BYOK API usage cost?
That will be similar to a rice grain. The cost will be so minimal that you will have to wait for a dollar to be fully spent ;).
Let’s do some calculations with OpenAI GPT-4o mini (recommended to use) at the current moment and rates.
## Cost Analysis for BYOK API Usage
To estimate the cost of using the BYOK API, let’s consider a sample of 1,000 minutes of video (equivalent to 50 videos, each 20 minutes long). Assuming an average of 150 words per minute, this equates to 150,000 total words.
OpenAI’s GPT-4o mini charges $0.150 per million input words. For 150,000 words, the input cost would be $0.0225. The output cost is 5% of the input, which in this case would be 7,500 words. At $0.600 per million output words, the cost would be $0.0045.
Therefore, the total cost for 1,000 minutes of video would be approximately $0.027.
Based on this calculation, the cost for 3,000 minutes would be $0.081, and for 10,000 minutes, it would be $0.27.
As AI models continue to become more affordable, BYOK is likely to be a cost-effective solution.
There will be different costs for different tasks, and translating summaries is one of them. Let’s see the calculation:
We are going to use the GPT-4o model for this one. We will translate an English summary into Spanish.
OpenAI’s GPT-4o charges $2.5 per million input words. For 1000 words, the input cost would be $0.00375. The output cost would be around $0.015.
So the total cost of translating a summary would be around $0.00375 (input) + $0.015 (output) = $0.01875.
Therefore, with $1, you can translate approximately 100,000 words of summaries into different languages using gpt4o.
We offer over 100 prompts in our library, so the average per prompt is around 200 words. Let’s assume you’re going to generate an 800-word blog post from a YouTube video transcript. So if we use GPT-4 mini here, then…
OpenAI’s GPT-4o mini charges $0.150 per million input words and $0.600 per million output words.
Input: Assuming 150 tokens per 100 words, 200 words would be approximately 300 tokens.
Output: Assuming a similar token-to-word ratio, 800 words would be approximately 1200 tokens.
Calculating Costs:
Input Cost:
300 tokens $0.150 / 1,000,000 tokens = $0.000045
Output Cost:
1200 tokens $0.600 / 1,000,000 tokens = $0.00072
Total Cost: $0.000045 (input) + $0.00072 (output) = $0.000765
Therefore, the estimated cost for processing 200 input words and generating 800 output words using GPT-4o mini is approximately $0.000765.
You can generate approximately 1389 blog posts with the same input and output for $1 using GPT-4o mini.
Note: You will only be charged if you use this feature.
Let’s assume we are going to play a 3-minute audio summary which will be around 1000 words. so let’s calculate the cost using gpt4o
Summary Length: 1000 words (estimated 1500 tokens)
Audio Generation Time: 3 minutes
Audio Token Count: 8000 tokens
Cost Calculation:
Input Cost:
1500 tokens $2.50 / 1,000,000 tokens = $0.00375
Output Cost:
8000 tokens $10.00 / 1,000,000 tokens = $0.08
Total Cost: $0.00375 (input) + $0.08 (output) = $0.08375
Therefore, using the revised estimate of 8000 tokens for the audio, the estimated cost for generating a 3-minute audio preview from a 1000-word summary is $0.08375.
Let’s do some calculations with OpenAI GPT-4o mini (recommended to use) at the current moment and rates.
## Cost Analysis for BYOK API UsageTo estimate the cost of using the BYOK API, let’s consider a sample of 1,000 minutes of video (equivalent to 50 videos, each 20 minutes long). Assuming an average of 150 words per minute, this equates to 150,000 total words.
OpenAI’s GPT-4o mini charges $0.150 per million input words. For 150,000 words, the input cost would be $0.0225. The output cost is 5% of the input, which in this case would be 7,500 words. At $0.600 per million output words, the cost would be $0.0045.
Therefore, the total cost for 1,000 minutes of video would be approximately $0.027.
Based on this calculation, the cost for 3,000 minutes would be $0.081, and for 10,000 minutes, it would be $0.27.
As AI models continue to become more affordable, BYOK is likely to be a cost-effective solution.
Translating Summary Cost
There will be different costs for different tasks, and translating summaries is one of them. Let’s see the calculation:
We are going to use the GPT-4o model for this one. We will translate an English summary into Spanish.
OpenAI’s GPT-4o charges $2.5 per million input words. For 1000 words, the input cost would be $0.00375. The output cost would be around $0.015.
So the total cost of translating a summary would be around $0.00375 (input) + $0.015 (output) = $0.01875.
Therefore, with $1, you can translate approximately 100,000 words of summaries into different languages using gpt4o.
Repurpose Content Cost
We offer over 100 prompts in our library, so the average per prompt is around 200 words. Let’s assume you’re going to generate an 800-word blog post from a YouTube video transcript. So if we use GPT-4 mini here, then…
OpenAI’s GPT-4o mini charges $0.150 per million input words and $0.600 per million output words.
Input: Assuming 150 tokens per 100 words, 200 words would be approximately 300 tokens.
Output: Assuming a similar token-to-word ratio, 800 words would be approximately 1200 tokens.
Calculating Costs:
Input Cost:
300 tokens $0.150 / 1,000,000 tokens = $0.000045
Output Cost:
1200 tokens $0.600 / 1,000,000 tokens = $0.00072
Total Cost: $0.000045 (input) + $0.00072 (output) = $0.000765
Therefore, the estimated cost for processing 200 input words and generating 800 output words using GPT-4o mini is approximately $0.000765.
You can generate approximately 1389 blog posts with the same input and output for $1 using GPT-4o mini.
Audio Playback Cost
Note: You will only be charged if you use this feature.
Let’s assume we are going to play a 3-minute audio summary which will be around 1000 words. so let’s calculate the cost using gpt4o
Summary Length: 1000 words (estimated 1500 tokens)
Audio Generation Time: 3 minutes
Audio Token Count: 8000 tokens
Cost Calculation:
Input Cost:
1500 tokens $2.50 / 1,000,000 tokens = $0.00375
Output Cost:
8000 tokens $10.00 / 1,000,000 tokens = $0.08
Total Cost: $0.00375 (input) + $0.08 (output) = $0.08375
Therefore, using the revised estimate of 8000 tokens for the audio, the estimated cost for generating a 3-minute audio preview from a 1000-word summary is $0.08375.
Updated on: 06/11/2024
Thank you!