# Agentics

> Unified AI Inference, Audio, Image Generation, and VPN Proxy Platform

## Website

- [Home](https://agentics.co.za/)
- [docs](https://agentics.co.za/docs)
- [models](https://agentics.co.za/models)
- [pricing](https://agentics.co.za/pricing)
- [products](https://agentics.co.za/products)
- [about](https://agentics.co.za/about)
- [contact](https://agentics.co.za/contact)

## API Reference

Base URL: https://api.agentics.co.za

### Authentication

All authenticated endpoints require a valid API key or JWT token
Header: `Authorization: Bearer YOUR_API_KEY`

### Authentication

User registration, login, and API key management

- POST `/v1/auth/register` - Register a new user
- POST `/v1/auth/login` - Login
- POST `/v1/auth/hardware-auth` - Hardware Authentication
- POST `/v1/auth/refresh` - Refresh Token
- GET `/v1/auth/profile` - Get Profile
- POST `/v1/auth/api-keys` - Create API Key
- GET `/v1/auth/api-keys` - List API Keys

### Inference

Text generation and chat completions using AI models

- GET `/v1/models` - List Models
- POST `/v1/generate` - Generate Text
- POST `/v1/chat/completions` - Chat Completions

### Images

AI image generation with multiple models and styles

- GET `/v1/images/models` - List Image Models
- POST `/v1/images/generations` - Generate Image

### Audio

Text-to-speech and speech-to-text services

- GET `/v1/audio/voices` - List Voices
- POST `/v1/audio/tts` - Text-to-Speech
- POST `/v1/audio/stt` - Speech-to-Text

### Live Transcription

Real-time audio transcription via WebSocket

- GET `/v1/audio/live/status` - Live Status
- WS `/v1/audio/live/ws` - Live Transcription WebSocket

### Agent

AI agent with tool execution, web search, and code execution capabilities

- GET `/v1/agent/tools` - List Tools
- POST `/v1/agent/message` - Agent Message
- POST `/v1/agent/search` - Web Search
- POST `/v1/agent/bash` - Execute Bash

### Realtime

Real-time conversational AI with integrated voice, similar to OpenAI Realtime API and Google Gemini Live. Features VAD (Voice Activity Detection), EOT (End-of-Turn detection), STT (Groq Whisper), TTS (Neural voices), and tool calling.

- GET `/v1/realtime/status` - Realtime Status
- WS `/v1/realtime` - Realtime Agent WebSocket

### Orb

Embeddable voice AI widget for websites. Secure session-based authentication allows embedding the Orb on third-party domains with your API key.

- POST `/v1/orb/session` - Create Orb Session
- GET `/v1/orb/embed` - Embed Orb Widget
- WS `/v1/orb/secure-ws` - Orb Secure WebSocket

### Embeddings

Text embeddings and vector search for semantic similarity

- POST `/v1/embeddings/generate` - Generate Embeddings
- POST `/v1/embeddings/add` - Add Embedding
- POST `/v1/embeddings/search` - Search Embeddings
- GET `/v1/embeddings/groups` - List Groups

### Research

Multi-engine web search, content extraction, and deep research with LLM-powered synthesis. Supports 7 search backends with automatic fallback.

- GET `/v1/research/status` - Research Status
- POST `/v1/research/search` - Web Search
- POST `/v1/research/contents` - Get Contents
- POST `/v1/research/findSimilar` - Find Similar
- POST `/v1/research/answer` - Quick Answer
- POST `/v1/research/deep` - Start Deep Research
- GET `/v1/research/deep/:taskId` - Poll Research Progress

### Glance

Fast in-process full-page text extraction and shallow same-origin crawler. Pulls title, headings, and body text from any URL. Use for quick reads, listings, article bodies, pricing and contact data, or anything that does not require a headless browser. Same engine powers the glance agent tool.

- GET `/v1/glance/status` - Glance Status
- POST `/v1/glance/scrape` - Structured Scrape
- POST `/v1/glance/text` - Flattened Text
- POST `/v1/glance/flat` - Flat Markdown

### VPN

VPN proxy service for routing traffic through global servers

- GET `/v1/vpn/servers` - List VPN Servers
- POST `/v1/vpn/setup` - Setup VPN
- POST `/v1/vpn/connect` - Connect to VPN
- GET `/v1/vpn/status` - VPN Status
- POST `/v1/vpn/disconnect` - Disconnect VPN Session

### Connectors

External service connectors. Email (IMAP/SMTP) is the first; calendar, messaging, storage and SQL follow. Each connector supports three autonomy tiers: draft only, review queue, and full autonomy.

- GET `/v1/connectors/email/providers` - Suggest Provider Settings
- POST `/v1/connectors/email/test` - Test Email Credentials
- POST `/v1/connectors/email` - Connect Email Account
- GET `/v1/connectors/email` - List Email Connectors
- GET `/v1/connectors/email/:id` - Get Email Connector
- PUT `/v1/connectors/email/:id` - Update Email Connector
- DELETE `/v1/connectors/email/:id` - Disconnect Email Account
- POST `/v1/connectors/email/:id/sync` - Force Sync
- GET `/v1/connectors/email/:id/messages` - List Recent Messages
- GET `/v1/connectors/email/:id/messages/:msgId` - Get Full Message
- POST `/v1/connectors/email/:id/send` - Send Email
- GET `/v1/connectors/email/:id/drafts` - List Drafts
- POST `/v1/connectors/email/:id/drafts` - Create Draft
- POST `/v1/connectors/email/:id/drafts/:draftId/approve` - Approve and Send Draft
- POST `/v1/connectors/email/:id/drafts/:draftId/reject` - Reject Draft
- DELETE `/v1/connectors/email/:id/drafts/:draftId` - Delete Draft

### Text Models

ALM, ALM-Max, ALM-Pro, AgentX, Bolt, Bolt-4, Bolt-4X, Bolt-Pro, Delta, Delta-Mini, Griffin, Magma, Merlin, Neo, Neo-Agent, Neo-Fast, Neo-Reason, Nitro, Omega, Orion, Orion-Light, Phoenix, Polaris, auto

### Image Models

AGImage, AgenPics, Anime, AnimeReal, Banana, Cinematic, DallE3, Epic, Fantasy, Flux, FluxFW, FluxKontext, FluxPro, FluxProRep, FluxRep, FluxSchnell, GPTImage, Ghibli, Ideogram, Imagen, ImagenRep, ImagenUltra, Juggernaut, LowPoly, Medieval, Pencil, Photo, Pixel, Playground, Portrait, Pro, Professional, Realism, Recraft, Retro, SD3Large, SDXL, SDXLRep, SDXLTurbo, Stock, ThreeD, Turbo, VanGogh, Watercolor, AgenPics, AGImage, Anime, AnimeReal, Banana, Cinematic, DallE3, Epic, Fantasy, Flux, FluxFW, FluxKontext, FluxPro, FluxProRep, FluxRep, FluxSchnell, GPTImage, Ghibli, Ideogram, Imagen, ImagenRep, ImagenUltra, Juggernaut, LowPoly, Medieval, Pencil, Photo, Pixel, Playground, Portrait, Pro, Professional, Realism, Recraft, Retro, SD3Large, SDXL, SDXLRep, SDXLTurbo, Stock, ThreeD, Turbo, VanGogh, Watercolor

## SDKs & Compatibility

The Agentics API is OpenAI-compatible. Point the official OpenAI, Anthropic or Google SDK at the base URL with an ag_ key:

```js
import OpenAI from 'openai';
const client = new OpenAI({ apiKey: 'ag_your_key', baseURL: 'https://api.agentics.co.za/v1' });
```

Published packages: `agentics` (CLI), `@agentics/taskman`.

## Pricing Tiers

- free R0/mo
- standard R99/mo
- growth R349/mo
- business R599/mo
- premium R849/mo
- enterprise R1995/mo

## Contact

- Website: https://agentics.co.za
- API: https://api.agentics.co.za
- Email: connor@agentics.co.za

## Optional

- [Full docs](https://agentics.co.za/llms-full.txt)
- [OpenAPI](https://agentics.co.za/openapi.json)
- [Live metadata](https://api.agentics.co.za/v1/docs/metadata)