Claude Code
Best for: Serious work on real repos
Anthropic's agent for your terminal, IDE and desktop. Reads the whole repo, runs commands and tests, and handles large multi-file changes well.
Visit Claude Code ↗Home / Use case
Explore a science or math question with AI
“I want to dig into a technical problem with help from AI.”
AI can search papers, run calculations and simulations, and suggest ideas to test. Treat outputs as hypotheses until checked by experiment or proof.
Fix it yourself in 3 steps
- State the question, what's already known, and how you'd know an answer is right.
- Ask the AI to find relevant papers and propose approaches, with code you can run.
- Check results independently (rerun, compare to known cases) before trusting them.
Copy this prompt
My question: [question]. Known so far: [background]. Find the most relevant papers with links, propose 3 approaches, and write runnable code for the most promising one. Test it on a case with a known answer first. Mark which results are verified and which are guesses. Stop after the first test.
Fill in the [brackets]. Each prompt says when the AI should stop, so it doesn't wander off.
Agents people used for this: Claude Code, Codex, Perplexity Comet, Gemini.
Inspired by real examples collected at agentusecases.ai.
📚 6 real stories of people doing this with AI agents
- Build data pipelines and analysis for a research paper · Claude Code
- Discover room-temperature magnetic semiconductor candidates with agents · custom-built or unnamed agent
- Explore open math problems with a multi-agent system · custom-built or unnamed agent
- Open recent protein-ligand structures in PDB · Perplexity Comet
- Repair formal proofs for a verified garbage collector · Codex
- Solve open math problems autonomously · Gemini
Collected at AgentUseCases.ai.
Check every citation yourself, strip student and personal names before pasting work in, and treat AI findings as leads to verify.
Tools for this
OpenAI Codex
Best for: Delegating tasks as PRs
OpenAI's coding agent, in the CLI and in the cloud. Good at handing off well-scoped tasks that come back as pull requests.
Visit OpenAI Codex ↗Claude
Best for: Writing, analysis, code
Anthropic's assistant: strong writing, analysis and coding, with projects and artifacts for longer work.
Visit Claude ↗ChatGPT
Best for: Widest feature set
OpenAI's assistant with the broadest feature set: voice, images, search and a large app ecosystem.
Visit ChatGPT ↗Grok Bot
Best for: Chores that repeat — done for you
The easiest AI helper for everyday people: show it a chore once — a morning brief, checking a bill, watching for a refund — and it keeps doing it while you're away, checking in only when it needs your OK. Works from the Grok app on your phone and computer (beta).
Visit Grok Bot ↗Grok
Best for: Real-time info, everywhere you go
xAI's assistant with real-time X data — on iPhone, Android, desktop, the web, X, and built into Tesla cars. Want it to do chores on its own? See Grok Bot.
Visit Grok ↗Gemini
Best for: Google Workspace users
Google's assistant, woven into Gmail, Docs and Android. Strongest if you live in Google's apps.
Visit Gemini ↗Perplexity
Best for: Research with citations
An answer engine that searches the web and cites its sources. Built for research questions.
Visit Perplexity ↗