Ensoria Capabilities

The complete inventory of all providers, models, APIs, and tools powering the Ensoria ecosystem — from cloud AI models to local inference, infrastructure, and membership plans.

Cloud AI Models

The primary and fallback models available via cloud providers, routed intelligently by the Ensoria Router.

ModelProviderUsageCost
DeepSeek V4 FlashDeepSeekPrimary model — main session agentIncluded (subscription)
DeepSeek V4 ProDeepSeekComplex reasoning, deep analysisIncluded (subscription)
OpenAI GPT-4oOpenAIEnterprise-grade reasoning, fallback€5–15/mo (API usage)
OpenAI GPT-4o MiniOpenAIFast, lightweight tasks€2–5/mo (API usage)
OpenAI GPT-4.1 seriesOpenAILatest GPT-4 generation€10–25/mo (API usage)
Google Gemini 2.5 FlashGoogleFast inference, cost-efficientFree tier + €5/mo
Google Gemini 2.5 ProGoogleComplex multimodal reasoning€10–20/mo
Google Gemini 3.1 Flash ImageGoogleImage analysis & generationFree tier + €3/mo
Perplexity SonarPerplexityWeb search, fact retrievalFree tier + €5/mo
Perplexity Sonar ProPerplexityAdvanced search with citations€10/mo
Perplexity Reasoning ProPerplexityMulti-step reasoning + deep search€15/mo
Perplexity Deep ResearchPerplexityIn-depth research reports€20/mo
OpenRouterOpenRouterAggregator — hundreds of modelsPay-as-you-go API

Local AI Models

On-device inference via Ollama on the M5 machine with 64GB RAM. Zero latency, zero API cost, full privacy.

ModelSizeQuantizationRAM Usage
Qwen 2.5 72B47 GBQ4_K_M~28 GB
Qwen 2.5 32B19 GBQ4_K_M~16 GB
Gemma 3 27B17 GBQ4_K_M~14 GB

All models run on M5 machine with 64 GB RAM via Ollama. Switchable on demand by Ensoria Router.

External Integrations

Every external service and API that Ensoria connects to, extending its capabilities beyond the model layer.

🔌 Google Workspace

Gmail, Drive, Contacts — OAuth 2.0 authenticated. Read & write access for email management, document retrieval, and contact synchronization.

🌐 Cloudflare

DNS management, Cloudflare Tunnel (4 QUIC replicas) for secure ingress without open ports.

🔧 MCP Server

24 tools, 3 resources. The Model Context Protocol bridge connecting Ensoria to external systems, databases, and services.

🎤 Whisper (Local STT)

OpenAI Whisper running locally via CLI. Zero API key required. Full privacy for voice transcripts.

🖼️ Image Analysis

Powered by Google Gemini. Analyze images, screenshots, diagrams, and visual content in conversations.

🎨 Image Generation

Powered by Google Gemini. Generate images from prompts directly within Ensoria.

🎬 Video Generation

Powered by Google Veo. Generate short video clips from text descriptions.

🎵 Music Generation

Powered by Google Lyria. Generate original music tracks from prompts.

🗣️ TTS

Text-to-speech for voice output. Supports multiple voices and languages.

🔍 Web Search

Powered by Perplexity and Google. Real-time web search integrated into Ensoria's reasoning flow.

📄 Web Fetch

Extract readable content (HTML → markdown/text) from any URL. Lightweight page access without browser automation.

Infrastructure Stack

The physical and virtual infrastructure that hosts Ensoria, from cloud VPS to network protocols.

🖥️ Hetzner VPS (Node 0)

Specs: 16 vCPU · 30 GB RAM · 610 GB NVMe + 1 TB HDD
Hosts the Ensoria Mother Node, back-end services, and persistent storage.

🔒 Cloudflare Tunnel

4 QUIC replicas
Zero-trust ingress, no open ports, DDoS protection, automatic failover.

🌐 DNAN Network Protocol

Distributed Neural Agent Network — P2P encrypted mesh for multi-node Ensoria deployments. Each peer is a sovereign agent node.

💻 OpenClaw on M5

Main session runs on AleX's Mac M5. Orchestrates all agents, tools, and model routing. Local inference via Ollama.

The Ensoria Ecosystem

Every module that makes up the Ensoria OS — from the chat agent to the management platform.

🤖 Ensoria Control

The main chat agent. Available on Telegram, WhatsApp, Signal, Discord, and WebChat. Your primary interaction point with Ensoria.

📊 Ensoria Manage

Full-stack management platform: CRM, Projects, Events, Requirements, Milestones. Web UI + REST API.

🧠 Ensoria Core

Brain (operational framework), Soul (values & principles), Router (model routing intelligence). The foundation of every Ensoria instance.

🌍 Ensoria Live

Website chat widget for live customer interactions. Embeddable in any site.

🔌 MCP Server

24 tools, 3 resources. The Model Context Protocol bridge connecting Ensoria to external systems and data sources.

By Membership Plan

Every resource mapped to the four membership tiers. Choose the plan that matches your needs.

Resource Personal Professional Startup Enterprise
AI Cloud ModelsDeepSeek + GeminiDeepSeek + Gemini + GPTDeepSeek + Gemini + GPT + PerplexityFull access + OpenAI
AI Local Models✅ Qwen 32B✅ Qwen 72B✅ Full local + custom
Image Analysis✅ Gemini✅ Gemini✅ Gemini✅ Gemini
Image Generation✅ Gemini✅ Gemini✅ Gemini✅ Gemini + custom
Video Generation✅ Veo✅ Veo + custom
Music Generation✅ Lyria
Google Workspace
MCP Server✅ (basic)✅ (full)✅ (full + custom)
Local Models✅ (32B)✅ (72B)✅ (full suite)
VPSCX32 (16GB)CX42 (32GB)CX52 (64GB)Custom
Storage100 GB250 GB500 GBCustom
Users11Up to 3Unlimited
ChannelsTelegram + WebChat+ WhatsApp+ Signal+ Discord + API
DNAN Node✅ Child Node✅ Full mesh
SupportCommunityPriorityPremiumDedicated
Price€20/mo€49/mo€99/moCustom

ROI by Plan

Estimated return on investment based on hours saved and typical professional billing rates.

Plan Cost/mo Hours saved/mo Value at €25/h Net ROI
Personal€20~20 h€500+€480/mo
Professional€49~40 h€1,000+€951/mo
Startup€99~80 h€2,000+€1,901/mo
EnterpriseCustom~160+ h€4,000+Custom

* Estimated at €25/h. Real ROI depends on your specific use case. For professionals billing €50–100/h, the ROI doubles or more.