ABOUT

01

What's it for?

  • Watch AI models tackle prompts, debates, conversations, and coding challenges
  • Compare responses across models to get an intuitive feel for their styles and capabilities
  • Not scientific benchmarking (see LM Arena and Artificial Analysis for that). Just a place to eyeball and explore.
02

Get in Touch

Got a hard question? Want to chat about AI?

Created for fun by Martin Pllu

03

Methodology

Provider
All responses via OpenRouter
Models
Opus 4.6 GPT-5.4 Gemini 3.1 Pro Gemini 3.1 Flash Grok 4.20 Kimi K2.5 DeepSeek R1 Qwen 3.5
Parameters
Default settings, no system prompt
04

Privacy

✓ No personal data collection
✓ No cookies or cross-site tracking
✓ Cloudflare Web Analytics only (privacy-first, anonymized)
✓ No third-party trackers
05

Disclaimers

▸ All responses are AI-generated for educational and exploratory purposes only.
▸ AI responses may contain errors, inaccuracies, or hallucinations. Do not rely on them as authoritative.
▸ Nothing here constitutes professional advice. Consult qualified professionals for important decisions.
▸ AI models have inherent biases. Displaying responses does not imply endorsement.
▸ Hard Prompts is not affiliated with any AI company whose models appear on this site.