Question 1

What is an AI API playground?

Accepted Answer

#1 AI API Playground and developer console, AI | API Console allows you to test, compare, and debug different artificial intelligence models from various providers using a single unified interface, without needing to write custom integration code first.

Question 2

How can I test an AI API?

Accepted Answer

Simply launch the AI | API Console playground, select your desired provider and model, securely enter your API key, configure the specific parameters you want to test (like temperature or max tokens), and send your prompt.

Question 3

Can I use my own API key?

Accepted Answer

Yes. Our platform uses a zero-retention local session architecture. You bring your own API keys, and they are only ever used to securely construct and dispatch requests directly from your browser to the provider.

Question 4

Can I test OpenAI APIs?

Accepted Answer

Yes, we support the entire modern OpenAI catalog, including their flagship GPT-4o models, o1 reasoning models, whisper transcription, and TTS endpoints.

Question 5

Can I test Claude APIs?

Accepted Answer

Yes, the console natively supports Anthropic's Claude 3.5 Sonnet and Opus models, allowing you to test their long-context windows and specific system prompts natively.

Question 6

Can I test Gemini APIs?

Accepted Answer

Yes, you can test Google Gemini 1.5 Pro and Flash directly, including their massive 1M+ token context windows and multimodal attachment capabilities.

Question 7

Can I compare AI models?

Accepted Answer

Absolutely. By switching models in the playground, you can run the exact same prompt across different providers and instantly compare response quality, inference latency, token usage, and API costs.

Question 8

Can I test image and audio inputs?

Accepted Answer

Yes, our console supports full multimodal testing. If a model supports it, you can seamlessly upload images, PDFs, CSVs, or audio files directly into the prompt box.

Question 9

Can I inspect raw JSON API responses?

Accepted Answer

Yes, every response in the console includes a toggle that allows you to instantly inspect the exact, raw, unparsed JSON payload returned by the provider for ultimate debugging clarity.

Question 10

Can I generate API code?

Accepted Answer

Yes. Using the live Code Builder, your current prompt and parameter configurations are instantly translated into working, copy-pasteable cURL, Python, Node.js, and Go scripts.

All Perplexity Models

Sonar (Chat)

Sonar Pro

Sonar Reasoning Pro

Agent Fast

Agent Low

Agent Medium

Agent High

Agent xHigh

Agent Wide Research