Whalli API

One OpenAI-compatible endpoint for every AI model on Whalli — chat, images, audio, music, and video. Point any OpenAI-compatible client or SDK at it and start building.

Why it works everywhere#

The API speaks the OpenAI wire format. Tools like OpenCode, Ollama-compatible clients, Cline, Cursor, LiteLLM, and the official OpenAI SDKs already know how to talk to it — you only change the base_url and your API key. No new SDK to learn.

Base URL#

url

https://api.studio.whalli.com/v1

Every endpoint on this page is relative to that base URL.

Quickstart#

1. Create an API key in the Console. 2. Export it. 3. Make your first request.

bash

export WHALLI_API_KEY="wh_live_..."

curl https://api.studio.whalli.com/v1/chat/completions \
  -H "Authorization: Bearer $WHALLI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.5",
    "messages": [{ "role": "user", "content": "Hello!" }]
  }'

Using the official OpenAI SDK? Only two settings change — base_url and api_key. Everything else is identical.

What you can do#

The API is organized into a few groups, each documented in its own page:

Chat completions
Text generation with streaming, vision, and reasoning.
Images
Generate images from a prompt.
Audio
Text-to-speech and speech-to-text.
Async jobs
Music, sound effects, and video — create then poll.

Conventions#

Requests and responses are JSON unless noted (audio endpoints return or accept binary). Authentication is a Bearer token. Usage is billed in credits against your Whalli account, and every request is rate-limited per key. These are covered in the guides.