C.W.K.
Stream
Lesson 03 of 04 · published

Decision Framework by Use Case

~10 min · compare, decision, use-cases

Level 0Scout
0 XP0/41 lessons0/12 achievements
0/100 XP to next level100 XP to go0% complete

Mapping use cases to paradigms

Below is a working map. None of these are absolute laws — but each row reflects a strong prior that should be the default unless you have a specific reason to deviate.

Use CaseBest paradigmWhy
Conversational chatbotDense (or MoE) standardPredictable cost, fast responses, good enough for conversation
Coding autocompleteSmall dense, standardLatency-critical; reasoning would feel like the IDE froze
Coding debugging assistantReasoning for hard, dense for autocompleteHard debugging benefits from extended thinking; trivial completions don't
Multi-step agent workflowMoE or small dense, standardMany calls × per-call latency = bad with reasoning; cost adds up too
Long-context retrievalDense or MoE, standardRetrieval is fast pattern matching, not deep reasoning
Local on-device inferenceSmall dense + quantizationBest ecosystem support, predictable memory, fits hardware budgets
Research/analysis (one-shot)ReasoningMulti-step analysis benefits from extended thinking
Creative writingStandard (any backbone)Reasoning often makes creative output more mechanical, not better
Math word problems / proofsReasoningWhere the reasoning paradigm shines most cleanly
Cost-sensitive production APIMoE standard or small denseMoE for capability/FLOP, small dense for predictability
Frontier-capacity researchMoE + reasoning if hardMaximum capability, accept the cost

How to reason past the table

For any new use case, ask three questions:

  1. Is this a chain of many calls (agentic) or a single hard call (analytic)? Many calls → avoid reasoning. Single hard call → consider reasoning.
  2. Is the bottleneck latency, cost, or quality? Latency → fast dense. Cost → small dense or MoE. Quality on hard problems → reasoning.
  3. Will this run locally or on a managed API? Locally → small dense. Managed → any.

The default

If you cannot articulate a specific reason to use anything else: medium dense (Llama 3.3 70B class) in standard inference mode, on a managed API. That is the productive default in 2026.

External links

Exercise

Take three different products you use that are LLM-backed (chat, code editor, search, summarizer, research tool, anything). For each one, infer which paradigm is most likely under the hood. Then think about whether that matches what the table would recommend. Where it doesn't, ask yourself why the product designers might have made that choice — and whether you'd make the same one.

Progress

Progress is local-only — sign in to sync across devices.
Spotted a bug or have feedback on this page?Report an Issue

Comments 0

🔔 Reply notifications (sign in)
Sign inPlease sign in to comment.

No comments yet — be the first.