Token counting and cost estimation in Prompt Engineering / GenAI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding token counting in text inputs

If a text input contains the sentence:
"Hello world! How are you today?"
How many tokens would a typical tokenizer count for this sentence?

A 8 tokens

B 6 tokens

C 5 tokens

D 7 tokens

💡 Hint

Think about how punctuation and spaces are usually split into tokens.

Most tokenizers split words and punctuation separately. Here, 'Hello', 'world', '!', 'How', 'are', 'you', 'today', '?' are 8 tokens.

❓ Predict Output

intermediate

Calculate token cost for a prompt

Given a prompt with 150 tokens and a model charging $0.002 per 1,000 tokens, what is the cost for processing this prompt?

A $0.03

B $0.0003

C $0.00015

D $0.003

💡 Hint

Calculate cost by multiplying tokens by price per token.

Cost = (150 tokens / 1000) * $0.002 = 0.15 * 0.002 = $0.0003

❓ Model Choice

advanced

Choosing a model based on token limits

You have a text input of 12,000 tokens. Which model should you choose if Model A supports up to 8,000 tokens and Model B supports up to 16,000 tokens?

A Either model works

B Model A

C Model B

D Neither model works

💡 Hint

Check the maximum token limit for each model.

Model A supports only 8,000 tokens, which is less than 12,000 tokens input. Model B supports 16,000 tokens, which is enough for 12,000 tokens input.

❓ Metrics

advanced

Estimating total cost for multiple API calls

You make 3 API calls with token usage: 500, 1,200, and 2,300 tokens. The cost is $0.0015 per 1,000 tokens. What is the total cost?

A $0.0075

B $0.009

C $0.0069

D $0.006

💡 Hint

Sum tokens first, then multiply by cost per token.

Total tokens = 500 + 1200 + 2300 = 4000 tokens. Cost = (4000 / 1000) * 0.0015 = 4 * 0.0015 = $0.006

🔧 Debug

expert

Identify the error in token cost calculation code

What error does this code produce?

tokens = 1200
price_per_1000 = 0.002
cost = tokens * price_per_1000
print(f"Cost: ${cost}")

Prompt Engineering / GenAI

tokens = 1200
price_per_1000 = 0.002
cost = tokens * price_per_1000
print(f"Cost: ${cost}")

A Prints Cost: $2.4

B Prints Cost: $0.0024

C SyntaxError due to f-string

D TypeError due to division

💡 Hint

Check the order of operations and units in the formula.

The code incorrectly omits the division by 1000. It calculates cost = tokens * price_per_1000 = 1200 * 0.002 = 2.4, printing Cost: $2.4, which is too high (correct would be $0.0024).