Experiment, benchmark, and tune AI models
The AI Playground gives you the environment to explore model behaviour, test system prompts, and validate tool integrations instantly.
Model mode
Test system prompts and model behaviour across dozens of models, with custom run settings.
Compare mode
Evaluate different models side-by-side under identical prompts, to benchmark output quality and accuracy.
Generative UI
Generate dynamic, personalised UIs from prompts, using pre-defined components designed by Cloudflare.
Voice agent · Coming soon
Test and customise voice assistants that listen, think, speak and transcribe in real time.
By using the AI Playground, you agree to our Terms and Privacy Policy. We use necessary cookies only. Models may have third-party licenses. Output is unverified and doesn't reflect Cloudflare's views.