With each analysis, you make a choice: fast and cheap or deep and comprehensive. This module clears up the confusion — it shows what changes (and what doesn’t), the models behind it, the costs side by side, and a simple rule for deciding.
Illustrative diagram — both modes produce the same documents; what changes is the depth of the research.
There’s a common myth that Comprehensive mode “generates more stuff.” That’s not true. There’s only one difference—the research depth. Quick runs about 9 searches; Comprehensive runs about 18. The 15 final documents are exactly the same.
Think of it like an essay: both modes write the same number of pages with the same structure. Comprehensive simply read twice as many sources before writing — so the pages are better supported, with more detail and nuance.
Quick is the default mode — you don't even need to type the flag. It's fast (2-3 minutes) and very inexpensive (~US$ 0,05). Think of it as your tool for screening.
# Estas duas linhas fazem a mesma coisa
python -m strategy_factory.main run "Stripe"
python -m strategy_factory.main run "Stripe" --mode quick
# ~9 buscas · 2-3 min · ~US$ 0,05For the overwhelming majority of cases, Quick already delivers a great draft. Move up to Comprehensive only when the result is really for external use (client, leadership).
Comprehensive is the in-depth pass. It takes 5 to 10 minutes and costs about US$ 0,50 — ten times more than Quick, but still less than a dollar. This is the mode for the company you're going to actually present.
# A flag é obrigatória aqui
python -m strategy_factory.main run "Stripe" --mode comprehensive
# Com contexto, o resultado fica ainda melhor
python -m strategy_factory.main run "Stripe" \
--mode comprehensive \
--context "B2B payments, fintech"
# ~18 buscas · 5-10 min · ~US$ 0,50"Ten times more expensive" sounds like a lot, but it’s the difference between 5 and 50 cents. When the result matters, it’s the best investment you’ll make that day.
Why does Comprehensive cost more? Because it uses more expensive and powerful AI models for research. Understanding which models are used in each mode explains the cost.
Uses only the model sonar from Perplexity — cheap and fast. It’s enough for a good overview.
Combines sonar, sonar-pro e sonar-deep-research. Stronger (and more expensive) models that do background research.
In both modes, the writing uses the same model: gemini-2.5-flash. That’s why the documents have the same format—the source material changes, not the writer.
The numbers vary a little depending on the company size and how much research is needed. Here are the actual ranges for both modes — and a reminder that generation is always free.
┌─────────────────┬────────────┬──────────────────┐
│ MODO │ TEMPO │ CUSTO │
├─────────────────┼────────────┼──────────────────┤
│ 🐇 Quick │ 2-3 min │ ~US$ 0,02 - 0,15 │
│ 🐢 Comprehensive│ 5-10 min │ ~US$ 0,31 - 0,90 │
│ ⚙️ Geração │ local │ GRÁTIS │
└─────────────────┴────────────┴──────────────────┘Generation (PPTX, DOCX, PNG) runs on your computer and doesn’t use the API—so it’s always free.
typical Quick cost
typical Comprehensive cost
generation phase
Want to know how much a specific analysis cost? Run status "Company" --detailed: it shows the cost recorded in state.json.
The winning strategy is simple: screen several companies with Quick and go deeper only on the one you choose with Comprehensive. And when you dig deeper, reuse the research you already have.
Você vai apresentar/entregar este resultado?
│
├─ NÃO → 🐇 Quick (triagem, estudo, primeira passada)
│
└─ SIM → 🐢 Comprehensive
│
└─ Já rodou Quick nesta empresa antes?
│
├─ SIM → --mode comprehensive --skip-research?
│ (avalie: pesquisa básica já serve?)
│
└─ NÃO → --mode comprehensive (pesquisa do zero)Run run "Company" for each candidate. It costs pennies in total.
Read the quick drafts and decide which company deserves an in-depth analysis.
Run --mode comprehensive on the selected one. Use --skip-research if basic research is enough.
Keep filtering cheap, and spend more only where it counts. This balance between Quick and Comprehensive keeps your costs tiny without sacrificing quality when it matters.
2.3 — Behind the scenes (concepts) — open the black box: the 3 phases behind the scenes, the state and cache files, and the output folder.