Try any model before you write a line of code.
Pick a model, set a system prompt, and chat. Every reply shows the tokens it used and what it cost, at the same price as the API.
Model
Create an account to send
Sign up, then add funds or choose a Token Plan. Playground replies are billed like any API request, with no gateway fees.
What you can do here
Every model, one place
Claude, GPT, Gemini, DeepSeek, GLM, and Kimi in one place. Start a new chat with another model and compare how they answer.
See the cost of every reply
Input, output, and cached tokens appear under each answer with the exact amount it was billed.
Text, images, and video
Chat with text models, or describe a picture or a scene and get images or a video clip back.
Popular models to try
Claude Fable 5.1Anthropic's most intelligent model, for the hardest problems.Try Claude Fable 5.1Claude Opus 5.5Long-running agentic coding at a lower price than Opus 5.Try Claude Opus 5.5GPT-6 AstraOpenAI's most capable model for the hardest end-to-end work.Try GPT-6 AstraGPT-5.6 SolFrontier GPT-5.6 reasoning for complex professional work.Try GPT-5.6 SolGemini 4 ArgonGoogle's frontier model for long-horizon coding and knowledge work.Try Gemini 4 ArgonGemini 3.8 FlashFast multimodal reasoning for coding and agents.Try Gemini 3.8 Flash