Generate-and-Rank
Reasoning
Level 2 · Advanced reasoningGenerate several candidate answers, score them, and return the best one — a simpler alternative to Tree of Thoughts.
Generate-and-rank asks the model to produce several candidate solutions, score each one against explicit criteria, and return the best-scoring one. It's a simplified relative of Tree of Thoughts: where Tree of Thoughts explores and evaluates branches as it develops them, generate-and-rank simply produces complete candidates first and compares them afterward — less structured, but much easier to set up and often good enough.
When to use it
Use it when you want more than one attempt without the full machinery of Tree of Thoughts — writing tasks with several valid directions, or any task where 'good' is somewhat subjective and comparing full candidates is more useful than exploring partial ones.
When to avoid it
Avoid it for tasks with a single correct answer (Self-Consistency is a better fit there), or when the criteria for 'best' are genuinely hard to state — if you can't describe what makes one candidate better than another, ranking them won't help.
Advantages
- Much simpler to set up than Tree of Thoughts, for a similar benefit on many tasks
- Makes the selection criteria explicit and visible, not just implicit in the model's first attempt
- Works well for subjective/creative tasks where Self-Consistency (majority voting) doesn't apply
Disadvantages
- Multiplies token cost by the number of candidates generated
- Only as good as the scoring criteria provided — vague criteria produce a vague ranking
- Doesn't explore or combine partial ideas across candidates the way Tree of Thoughts can
Examples
Write a tagline for a productivity app aimed at freelancers. Generate 5 candidate taglines. Score each from 1-5 against these criteria: memorability, clarity about the product, and appeal to freelancers specifically. Return the highest-scoring one, with its score and a one-line justification.
Pairs well with
Try it
Fill in the placeholders below and see the compiled prompt update live.
What you want to achieve.
How many candidate solutions to generate. Default: 5.
How should the alternatives be compared? (e.g. cost, impact, time to ship)
- 24
- Words
- 196
- Characters
- 0/3
- Placeholders filled
- 1 min
- Reading time
- Completeness0
- Clarity0
- Context0
- Specificity0
- Constraints0
- Format0
- Output0
- Constraints note0
- Success criteria0
- Scope focus100
- Verification step0
- Negative example0
- Confidence0
- Placeholder syntax100
- Markdown structure0
- Fill in the remaining placeholders.
- Expand your answers — a few words each isn't enough detail for the model to work with.
- Add background context so the model understands the situation.