GenAI Field GuideGetting started
Which tool for which job
Pick the tool by the shape of the task, load it with the right document, and verify the result. That combination beats "the smartest model" reasoning from memory.
GenAI.mil-first cheat sheet
| Task | First pick | Why |
|---|---|---|
| Summarize or interrogate a 200-page PDF, a slide deck, a scanned form, a video | Gemini | Context size; native document, image, and video handling |
| Draft, rewrite, iterate text: memos, SOPs, evaluations, talking points | ChatGPT Mil (or Gemini) | DoD's own guidance; mature Projects and custom-GPT workflow |
| Multi-step research with lots of searching, or "keep working on this until done" | Grok, then verify hard | Multi-step agent tooling; but verify, given well-documented past incidents |
| Build a reusable helper for your shop (an "agent") | Gemini Agent Designer | The only built-in builder on GenAI.mil today |
| Write or debug code (if permitted in your role) | ChatGPT Mil / Codex-style tooling; Gemini | Strongest software workflows |
| Anything classified | Nothing on GenAI.mil at IL5 | Wrong impact level. Full stop. |