Public preview
Choosing the Model by Task
Don't default to the biggest model. Choose by: task depth, cost, privacy, speed.
Goal
Save time and cost by matching model to task instead of using Opus/GPT-4 for everything.
Steps
-
1
Quick simple task (classify, rephrase) → small fast model (Haiku, gpt-4o-mini).
Expected Outcome
Immediate
Save 30-70% of API cost without quality loss.
Evidence base
School: Applied LLM Engineering
Founders: Anthropic Research Team · Andrej Karpathy (2024)
Anthropic's model selection guide and Karpathy's general advice recommend starting smaller and upgrading only when needed.
Keywords
Frequently asked questions
How long does "Choosing the Model by Task" take?
This exercise takes about 10 minutes, practiced solo in a Written reflection format.
When will I notice the effect of "Choosing the Model by Task"?
Save 30-70% of API cost without quality loss. In the short term: Your daily responses become faster because you use lighter models.
Is "Choosing the Model by Task" evidence-based?
Yes — it draws on Applied LLM Engineering (Anthropic Research Team, Andrej Karpathy (2024)). Anthropic's model selection guide and Karpathy's general advice recommend starting smaller and upgrading only when needed.
Is "Choosing the Model by Task" suitable for beginners?
Its difficulty level is: Intermediate. No external tools required — it can be practiced directly inside the app.
Subscribe for full access
Subscribe to Plus for full access to this practice and the whole catalogue, or try first — free, no sign-in, no commitment.
7 complete practices (one per domain) are free forever, no sign-in.