What drives inference costs in Amazon Bedrock LLMs?

AIF-C01 By islever Editorial Last updated Aug 19, 2026

Short answer

A Number of tokens consumed

Question

A company wants to assess the costs that are associated with using a large language model (LLM) to generate inferences. The company wants to use Amazon Bedrock to build generative AI applications. Which factor will drive the inference costs?

Number of tokens consumed Correct Answer
Temperature value
Amount of data used to train the LLM
Total training time

Key Takeaways

The question tests the distinction between inference pricing models and training cost factors, highlighting that users pay for what they process (tokens), not how the model was built.

Inference costs in Amazon Bedrock are primarily driven by the number of tokens consumed during input and output processing. The community consensus confirms that token volume, not training metrics or model parameters like temperature, determines pricing.

Common mistake: Learners often confuse Option C (Amount of data used to train) or Option D (Total training time) with operational costs, failing to recognize that these are sunk costs associated with model development rather than per-request inference fees.

Expert Analysis

Why the Answer Is Correct

Option A is correct because Amazon Bedrock, like most generative AI services, charges based on the volume of tokens processed. Tokens represent the basic units of text (words, subwords, or characters) that the model reads as input and generates as output. Since computational resources are directly proportional to the amount of data processed, the number of tokens is the primary driver of inference costs.

Why the Other Options Are Wrong

Options C and D relate to the pre-training phase, which involves building the model before it is available for public use; companies do not pay per-inference for the historical training data or time. Option B (Temperature) is a hyperparameter that controls the randomness of the output but does not affect the computational load or the number of tokens generated, thus having no impact on cost.

Community Comment Notes

Commenters unanimously agree that tokens are the fundamental unit of billing in generative AI. Several comments emphasize that while training costs exist, they are separate from the operational expenditure (OpEx) of running inferences. The consensus reinforces that input and output token counts are the sole metrics for calculating inference bills.

Community Votes

100%

100% of anonymous learners picked answer A. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 4 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 4

PHD_CHENG 👍 3 Answered A

A is correct. Token is the basic unit of generative AI model
OnePG 👍 2 Answered A

A. Number of tokens consumed. More tokens used = higher cost. All other affects training costs, not inference costs. Correct answer is A
Jessiii 👍 1 Answered A

In the context of using Amazon Bedrock and generative AI models, inference costs are typically driven by the number of tokens consumed during the input and output processing. Number of tokens consumed refers to how many tokens (words, subwords, characters) the model processes during inference (both input and output). More tokens mean higher processing and hence higher costs.
85b5b55 👍 1 Answered A

No. of tokens consumed while processing. Tokens are the basic units of input and output that a generative AI model operates on, representing words, subwords, or other linguistic units.

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Practice All AIF-C01 Questions

Access 100 questions with complete answers and detailed explanations.

View Full AIF-C01 Practice Test →

← Back to AIF-C01 Study Guide