Perplexity Deep Search

**Judgement:** Perplexity Deep Search is a solid, focused wrapper around the Perplexity API that provides three well-defined search depths (search, reason, research) and practical CLI ergonomics. It is best suited for workflows that need fresh, web-grounded information or structured research directly from scripts. **Strengths & Fit:** - Good for quick factual lookups, multi-step reasoning, and long-form research from the same interface. - Helpful options for recency, domain filtering, language selection, and JSON output make it integration-friendly. - Clear setup and usage examples lower the barrier to adoption for users comfortable with shell scripts. **Key Risks & Limitations:** - Hard dependency on the external Perplexity API: outages, latency, rate limits, or API changes will directly impact functionality. - All query data is sent to a third-party service; this may be unsuitable for sensitive, proprietary, or regulated information. - Cost can become significant in research mode, which is explicitly described as high-cost and multi-call. - Result quality and safety are constrained by Perplexity’s models, web sources, and any associated biases or occasional inaccuracies. **Recommended Scenarios:** - Automating up-to-date web-informed research (news tracking, market scans, technical trend reports) where external data sharing is acceptable. - Building internal tooling that needs structured Perplexity outputs (via `--json`) for downstream analysis or dashboards. - Developer and analyst workflows that frequently compare technologies, summarize documents, or perform domain-constrained literature scans. **Less Suitable For:** - Highly sensitive, confidential, or regulated data queries. - Environments requiring strict offline operation or no reliance on third-party APIs. - Use cases where costs of repeated deep research queries are tightly constrained and must be predictable.