Quickstart

https://omniquasar.com/api/proxy/v1

curl -X POST https://omniquasar.com/api/proxy/v1/chat/completions \
  -H "Authorization: Bearer omni_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "omni-fast",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Authentication

All API requests require an API key passed as a Bearer token in the Authorization header.

Authorization: Bearer omni_your_key_here

• API keys are prefixed with omni_
• Create, rotate, and revoke keys from the Dashboard → API Keys page
• Never share or commit your API key to source control

Models

Choose a model based on your quality and cost requirements.

Credits & Billing

Credits are deducted based on input token count only. Output tokens are free.

Formula: credits deducted = base_credits × model_multiplier

Tiered deduction by input context size

Plans

Extra credits: $5 = 150 credits · one-time purchase · never expire

API Reference

POST https://omniquasar.com/api/proxy/v1/chat/completions

Request Parameters

Non-streaming example

curl -X POST https://omniquasar.com/api/proxy/v1/chat/completions \
  -H "Authorization: Bearer omni_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "omni-fast",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

{
  "id": "chatcmpl-xxx",
  "model": "omni-fast",
  "choices": [{
    "message": { "role": "assistant", "content": "Hello! How can I help?" },
    "finish_reason": "stop"
  }],
  "usage": { "prompt_tokens": 10, "completion_tokens": 12 }
}

Streaming example

curl -X POST https://omniquasar.com/api/proxy/v1/chat/completions \
  -H "Authorization: Bearer omni_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"model":"omni-fast","messages":[{"role":"user","content":"Hello!"}],"stream":true}'

data: {"choices":[{"delta":{"content":"Hello"}}]}
data: {"choices":[{"delta":{"content":"!"}}]}
data: [DONE]

Note: Omni API is powered by a stack of Claude models working together. When streaming is enabled, the connection is kept alive with SSE heartbeats during processing. The response content arrives after processing completes — not token-by-token. This is fully compatible with Claude Code, Roo Code, and Kilo Code.