Build with Sovereign AI
Integrate sovereign AI inference into your applications. SSE streaming, zero external dependencies, OpenAI-compatible API surface.
Quick Start
Choose a Model
Select from Edge models (free, sovereign) or Luxury tier (GPU-accelerated, admin-locked).
Send a Request
POST to /api/v1/sovereign/generate with your prompt and model ID.
Stream the Response
Enable stream: true for real-time SSE token streaming with <100ms first-token latency.
Code Examples
# Standard generation
curl -X POST https://sovereign.drpcoa.com/api/v1/sovereign/generate \
-H "Content-Type: application/json" \
-d '{
"model_id": "s-sonic-edge-nexus",
"prompt": "Explain quantum computing in simple terms",
"parameters": {
"max_tokens": 512,
"temperature": 0.7,
"stream": false
}
}'
# SSE Streaming (recommended)
curl -N -X POST https://sovereign.drpcoa.com/api/v1/sovereign/generate \
-H "Content-Type: application/json" \
-d '{
"model_id": "s-sonic-edge-nexus",
"prompt": "Write a Python fibonacci function",
"parameters": { "stream": true }
}'API Endpoints
Generate text, code, or structured output from any Edge model. Supports SSE token streaming.
| Parameter | Type | Required | Description |
|---|---|---|---|
| model_id | string | required | Model identifier (e.g. s-sonic-edge-nexus) |
| prompt | string | required | Input prompt for generation |
| parameters.stream | boolean | optional | Enable SSE token streaming (recommended) |
| parameters.max_tokens | number | optional | Maximum tokens to generate (default: 2048) |
| parameters.temperature | number | optional | Sampling temperature 0-2 (default: 0.7) |
Available Models
| Model ID | Name | Engine | Type | Tier | Status |
|---|---|---|---|---|---|
| s-sonic-edge-nexus | Edge Nexus | S-Sonic Edge | Text/Code | Free+ | Active |
| s-sonic-edge-echo | Edge Echo | S-Sonic Edge | Voice | Edge+ | Configured |
| s-sonic-edge-vision | Edge Vision | S-Sonic Edge | Image | Edge+ | Configured |
| s-sonic-edge-motion | Edge Motion | S-Sonic Edge | Video | Edge+ | Configured |
| s-sonic-v5-text | V5 Architect | S-Sonic V5 | Text | Luxury | Circuit Breaker |
| s-sonic-v5-flow | V5 Flow | S-Sonic V5 | Text | Luxury | Circuit Breaker |
| s-sonic-v5-physics | V5 Physics | S-Sonic V5 | Text | Luxury | Circuit Breaker |
Architecture
Circuit Breaker
All requests route through the Circuit Breaker. S-Sonic V5 Luxury tier is disabled by default — all inference runs on sovereign S-Sonic infrastructure at zero cost. Admin can toggle S-Sonic V5 via the SSONIC_V5_ENABLED environment variable.
SSE Token Streaming
Set stream: true to receive tokens via Server-Sent Events (SSE). Each chunk is a JSON object with a content field. The stream terminates with data: [DONE]. First-token latency <100ms on Nexus.
Authentication
Sovereign uses DrPCoA Passport for SSO authentication. Public endpoints (text, docs, suite) are accessible without auth. Protected endpoints require a valid session-token cookie from Passport login.
Infrastructure
Sovereign AI S-Sonic Series with 500B+ combined parameters (Q4_K_M quantization). S-Sonic Edge Vision & Motion headless for image/video generation. FastAPI for voice inference. All containerized via Docker Compose.
Response Formats
{
"success": true,
"data": {
"content": "Generated text here...",
"model": {
"id": "s-sonic-edge-nexus",
"name": "S-Sonic Edge Nexus",
"engine": "ollama"
},
"metadata": {
"circuit_breaker": "s-sonic",
"cost_usd": 0,
"tokens_used": 142,
"latency_ms": 1832
}
}
}data: {"content":"Hello"}
data: {"content":" there"}
data: {"content":"!"}
data: {"content":" How"}
data: {"content":" can"}
data: {"content":" I"}
data: {"content":" help"}
data: {"content":" you"}
data: {"content":"?"}
data: [DONE]