AI Model Decision Wizard
A guided flow that recommends AI models with explanations and trade-offs rather than a simple ranking. Answer 6 quick questions about your use case, budget, latency, language, privacy, and multimodal needs.
Decision Wizard
What is your primary use case?
What is your monthly budget?
What is your latency requirement?
Do you need multimodal input?
Do you need to self-host or avoid vendor lock-in?
What is your expected monthly volume?
How the Wizard Works
The wizard collects your answers across 6 dimensions — use case, budget, latency, multimodal needs, privacy, and volume — and scores every model in the catalog. The top 3 matches are presented with explanations of why each model fits and what trade-offs you're making. Unlike a simple ranking, this helps you understand the reasoning behind each recommendation.
FAQ
How is the recommendation score calculated?
Each model starts at 50 points. Points are added for matching your use case (+20), required capabilities (vision +10, reasoning +12, tools +8, etc.), budget fit (up to +15 for cheap models), latency fit (+8 for economy/balanced), and privacy needs (+6 for open weights). The top 3 scores are shown.
Why isn't the cheapest model always recommended?
Because quality matters. If your use case is coding or research, an economy model may produce lower-quality output that costs more in human review time. The wizard balances cost against quality fit for your specific use case.
Can I change my answers?
Yes. Use the Back button to revisit previous steps, or simply restart by refreshing the page.