Claude vs Codex
Using only the use cases on agentsetups: 168 name Claude, 69 name Codex, and 42 name both. Claude’s biggest category here is Agent setup & config (26); Codex’s is Coding (14).
This compares what people used each agent for in the sources we list, not which is better. Counts reflect what gets posted, not market share.
Use cases by category
| Category | Claude | Codex |
|---|---|---|
| Coding | 24 | 14 |
| Agent setup & config | 26 | 13 |
| Games | 21 | 6 |
| Content & video | 12 | 6 |
| Productivity & ops | 12 | 3 |
| Research & science | 9 | 3 |
| Design & frontend | 9 | 3 |
| Knowledge & memory | 8 | 3 |
| Hardware & IoT | 5 | 3 |
| Marketing & growth | 8 | 1 |
| Finance & trading | 5 | 2 |
| Personal life | 5 | 2 |
| Automation & integrations | 5 | 2 |
| Security | 4 | 2 |
| Personal assistants | 2 | 2 |
| DevOps | 4 | 1 |
| Sales & outreach | 2 | 1 |
| Education | 2 | 1 |
| Data analysis | 4 | 0 |
| Travel | 0 | 1 |
| Health | 1 | 0 |
Used together 42
- Find an exoplanet candidate in NASA TESS data
Claude Code ran 74 analyses and 1,000+ scripts; hidden-year transit predictions matched 3/3; Codex reviewed.
- Plan with Claude Code, execute with Codex via MCP
Claude Code plans and reviews, Codex executes through the Codex MCP, and notes and skills live in Obsidian.
- Decompile a full game with a long-running agent swarm
Over 3 months and 500B tokens, worker + reviewer agents reached 99% of functions, 83% byte-exact.
- Build paid web apps with an Opus front end and Codex
The founder’s playbook behind $100K MRR: one-shot a baseline, name the stack, use a quality reference per feature, self-score.
- Give one prompt and 6 hours for a 3D visualisation
Same prompt to GPT-6 Astra in Codex and Opus 5.5 in Claude Code; Opus ran 6 subagents and finished a three.js piece in 1h25.
- Let your different agents talk to each other over Tailscale
A personal tool let Grok Bot ask Muse to make a phone call; it connects Instinct, Codex, Claude Code and Hermes.
Different use cases
Only Claude 126
- Run agent teams on materials-science discovery · Research & science
- Build and backtest a trading bot with no coding background · Finance & trading
- Run Sonnet as main, Haiku explorers and Opus as advisor · Agent setup & config
- Give agents documentation instead of memory plugins · Knowledge & memory
- Install a skill that stops agents burying the answer · Agent setup & config
- Build a strategy backtesting engine with Claude Code · Finance & trading
- Run coding agents like a hospital: Db2 support in 2 days · Coding
- Make your game legible to AI agents · Games
Only Codex 27
- Have your agent plan and book a whole trip · Travel
- Produce 5-minute AI video ads from a Codex pipeline · Content & video
- Replace a pro app subscription with your own build · Coding
- Have Codex build a drop-in renderer to fix perf · Coding
- Build a hand-controlled game projected on your wall · Games
- Rebuild your personal site with Codex and a deploy tool · Design & frontend
- Let Codex operate a real iPhone over USB · Hardware & IoT
- Run OpenClaw on a local Qwen model at 150–250 tok/s · Agent setup & config