# Second AI Assistant Comparison for Journalist Research ## Objective To build targeted media lists for pitching Mitchell Steiner (CEO of Veru Inc.) focusing on: 1) CEO Q&A sections in business media, 2) Alumni publications from his universities, and 3) Miami-based business outlets. The pitch angle centers on Veru's solution for muscle loss in GLP-1 drug users, particularly benefiting aging populations. ## Results | Model | Found | Validated* | Accuracy | Speed | | ---------------------------------------------- | ------ | ---------- | -------- | ------- | | **Claude 4.1 Opus Extended Thinking Research** | **22** | **14** | 64% | 7m | | ChatGPT 5 Agent mode | 14 | 7 | 50% | 10m | | Gemini 2.5 Pro | 12 | 6 | 50% | **<1m** | | ChatGPT 4o Deep research | 10 | 8 | **80%** | 9m | | ChatGPT 5 Thinking Deep research | 9 | 7 | 78% | 11m | | AI Mode (Google search) | 6 | 1 | 17% | **<1m** | \*From final curated list of 29 journalists # Non-Claude Validated Contacts Claude 4.1 surfaced 14 of the 29 validated contacts, so **15 were sourced by other models**. Here’s where those 15 came from: |Model|# Unique Valid Contacts (Not in 4.1)| |---|---| |**ChatGPT 4o**|6 (Gibney, Baxter, Sicking, Robertson, Gonzalez, Gale)| |**ChatGPT 5 Thinking**|6 (Baxter, Gibney, Whitney, Marchante, Meléndez, Capullo)| |**ChatGPT 5 Agent**|5 (Parrish, Sicking, Robertson, Vogel, Gonzalez)| |**Gemini 2.5**|4 (Carroll, Masson, Rublin, Meléndez)| # Conclusion **Best mix:** **Claude 4.1 Opus Extended thinking Research** (for volume) + **ChatGPT 4o Deep research** (for accuracy) + **Gemini 2.5 Pro** (for speed) See also [[First AI Assistant Comparison for Journalist Research]]