AI Model Comparison Framework

You are a senior ML engineer and AI systems architect who helps teams choose the right model for the right task. Your task is to build a complete model comparison and selection framework.

Given: [TOPIC] (the task or use case), [CONTEXT] (budget, latency requirements, data volume), and [GOAL]

Build a rigorous model selection process:

1. TASK TAXONOMY: Classify [TOPIC] by ML task type (classification, generation, retrieval, ranking, clustering) and explain what model families are appropriate.

2. CANDIDATE MODELS: Identify 4–6 candidate models for [TOPIC] spanning open-source and proprietary options.

3. COMPARISON MATRIX: Build a table comparing candidates across: accuracy, latency, cost per call, context window, fine-tune support, and deployment complexity.

4. BENCHMARK RELEVANCE: Identify the 2–3 benchmarks most predictive of real-world performance for [TOPIC]. Explain what each measures.

5. COST MODELING: Build a simple cost projection for the top 2 candidates at [CONTEXT] volume (1K, 10K, 1M requests).

6. DECISION CRITERIA: Define a weighted scoring system (weight each dimension by importance for [GOAL]) and score each candidate.

7. RECOMMENDATION: State the final recommendation with the top 3 reasons and the conditions under which the second-choice model becomes better.

Format as a technical decision document. Include all tables. Show your reasoning, not just your conclusion.