# RandomBench: Ask an AI for a Random Number. It Says 7.

> Revnu is your AI growth hire. It finds customers, runs email and LinkedIn outreach, manages paid ads, and publishes content for Google and AI search.

Canonical: https://revnu.com/randombench

We asked five frontier models to pick a random number between 1 and 10, 100 times each. 488 of 500 answers were 7.

The same habit decides what a business tries next. On NoveltyBench, an independent benchmark from researchers at Carnegie Mellon, Revnu gives 9.6 different answers in 10 where the best frontier model gives 3.1, and scores 7.5 against 3.4 for answers that are both new and good. Asked for a stalled business's next growth test, Revnu produced 6.6 different good ideas in 20 asks against 2.5 for asking a frontier model directly.

## Machine-readable resources

- Agent setup recipe ("Set me up with revnu.com"): https://revnu.com/setup.md
- Revnu for AI agents: https://revnu.com/agents.md
- AI catalog (ARD): https://revnu.com/.well-known/ard.json
- API catalog (RFC 9727): https://revnu.com/.well-known/api-catalog
- MCP server card: https://revnu.com/.well-known/mcp/server-card.json
- Agent install script (MCP server + Revnu skill): https://revnu.com/install
- LLM index: https://revnu.com/llms.txt
- Full LLM context: https://revnu.com/llms-full.txt
- Blog index: https://revnu.com/blog/llms.txt
- Comparisons index: https://revnu.com/comparisons/llms.txt
- Use cases index: https://revnu.com/use-cases/llms.txt
- Guides index: https://revnu.com/guides/llms.txt
- Docs index (how to use Revnu, plus the API, CLI and MCP): https://revnu.com/docs/llms.txt
- Every docs page in one file: https://revnu.com/docs/llms-full.txt
- OpenAPI: https://revnu.com/openapi.json
- Sitemap: https://revnu.com/sitemap.xml
- Robots: https://revnu.com/robots.txt