Where ideas meet every model.
Chat, compare and refine prompts across the whole catalogue. When it works, export it to code with one click.
Everything at your fingertips
Side-by-side compare
Run one prompt across several models and see the differences instantly.
Fine control
Temperature, length, system prompts and more — saved as reusable presets.
Memory on tap
Toggle memory to keep context across conversations.
Skills in chat
Let the model search, read files or run code mid-conversation.
Export to code
Turn any conversation into a ready-to-run API request.
Share with your team
Share threads and presets so great prompts spread fast.
Fast enough to think with
Responses stream token by token with low latency, so iteration feels like conversation, not waiting.
- Streaming responses by default
- Markdown, tables and code rendered beautifully
- Keyboard-first, mobile-friendly
Explore, refine, ship
- 01
Explore
Try your idea on several models at once.
- 02
Refine
Tune prompts and settings, save presets.
- 03
Ground
Add memory and skills for real-world context.
- 04
Ship
Export to code and go live through the API.
Playground questions
Is the playground free?
Yes. The Free plan includes 2M tokens every month, usable in the playground, agents and API.
Does it work on mobile?
Yes. The playground is fully responsive and installable as an app on your phone.
Are my conversations private?
Yes. Conversations belong to your workspace and are never used to train shared models.
Can I export to my language of choice?
Export produces a standard HTTP request you can use from any language.