[LLM] When should I use the standard model or reasoning model?

You should use reasoning models or standard models depending on the complexity and type of task:

Use a Reasoning Model When:

• The task requires complex, multi-step reasoning, such as solving puzzles, advanced math problems, or logical challenges.
• You need to break down problems into smaller steps or provide explanations for answers, such as in coding, scientific analysis, or ambiguous queries requiring inference.
• Tasks involve intricate analysis or decision-making that goes beyond straightforward retrieval or synthesis.

Use a Standard Model When:

• The task is simple and straightforward, such as summarization, translation, or basic question-answering.
• You need efficiency in terms of cost and speed, as reasoning models are more resource-intensive and verbose.
• Tasks focus on retrieval and synthesis without requiring deep logical analysis (e.g., Retrieval-Augmented Generation).

In short, use reasoning models for complex problem-solving and standard models for simpler tasks to optimize performance and cost.