VitaAvanza Intelligence Layer — v2.1.0

The AI Models.

VitaAvanza develops purpose-built AI models trained on the intersection of human behavior, institutional systems, and real-life decision-making. Not general intelligence. Specific, deep, human intelligence.

vitaavanza.com/intelligence/models v2.1.0 · eu-west
$vita models --list --verbose --region=eu-west
 
>Authenticating with VitaAvanza Intelligence...
> Connected · Latency: 12ms · Region: EU-West-1
 
>Scanning model registry...
> 3 models found · 3 active · 0 deprecated
 
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
MODEL TIER STATUS CONTEXT
─────────────────── ──────── ────────── ─────────
mitra-1.3g-mini free active 128K
mitra-1.2 pro active 200K
vareon-1 business active 512K
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 
>All models operational. DVI layer: online.
$
scroll to explore
0
Active models
0
Avg response
0
Life situations analyzed
0%
Uptime SLA
03 · Intelligence Visualization
vareon-1.core / 6 satellite nodes / orbital reasoning
live · stream · eu-west-1
04
Model catalog
three tiers · one philosophy
Entry Level

Mitra

1.3G Mini · vitaavanza/mitra-mini

Fast, efficient, and capable of handling everyday tasks — from website generation to basic life guidance. Optimized for speed and accessibility. The right tool for most everyday needs.

speed Fast
reasoning Standard
context 128K
multilingual 14 langs
Website gen Life guidance Doc Q&A Task plan Translate
Tier Free
Active
◈ Flagship

Vareon

1 · vitaavanza/vareon-flagship

VitaAvanza's most powerful model. Designed for institutional-grade reasoning — detecting patterns across populations, generating investor-ready outputs, and understanding the full complexity of human systems at scale. Where precision is not optional.

speed Quality
reasoning Institutional
context 512K
multilingual 42 langs
Population analysis DVI computation Risk detection Institutional reports Life OS reasoning Pitch generation Crisis prediction
Tier Business
Active
◆ Advanced

Mitra

1.2 · vitaavanza/mitra-advanced

Deeper reasoning, stronger design intelligence, and more nuanced understanding of complex life situations. Built for founders, professionals, and institutions that need precision and depth beyond everyday tasks.

speed Balanced
reasoning Deep
context 200K
multilingual 22 langs
Advanced design Life analysis Investor mode Positioning Brutal feedback Brand intel
Tier Pro
Active
05
API reference
streaming · openai-compatible
POST api.vitaavanza.com/v1/models/completions
curl -X POST https://api.vitaavanza.com/v1/models/completions \
  -H "Authorization: Bearer va_live_••••••••••••" \
  -H "Content-Type: application/json" \
  -H "X-VitaAvanza-Version: 2024-01" \
  -d '{
    "model": "vareon-1",
    "messages": [
      {
        "role": "user",
        "content": "Analyze this life situation..."
      }
    ],
    "max_tokens": 4096,
    "stream": true,
    "context": {
      "dvi_profile": "user_12345",
      "institution_id": "uni_trento_01"
    }
  }'
from vitaavanza import VitaAvanza

client = VitaAvanza(api_key=os.environ["VA_API_KEY"])

stream = client.models.completions(
    model="vareon-1",
    messages=[
        {"role": "user",
         "content": "Analyze this life situation..."}
    ],
    max_tokens=4096,
    stream=True,
    context={
        "dvi_profile": "user_12345",
        "institution_id": "uni_trento_01",
    },
)

for chunk in stream:
    print(chunk.delta.content, end="")
import { VitaAvanza } from "@vitaavanza/sdk";

const client = new VitaAvanza({
  apiKey: process.env.VA_API_KEY,
});

const stream = await client.models.completions({
  model: "vareon-1",
  messages: [
    { role: "user",
      content: "Analyze this life situation..." },
  ],
  max_tokens: 4096,
  stream: true,
  context: {
    dvi_profile: "user_12345",
    institution_id: "uni_trento_01",
  },
});

for await (const chunk of stream) {
  process.stdout.write(chunk.delta.content);
}
package main

import (
    "context"
    "fmt"
    "os"

    va "github.com/vitaavanza/vitaavanza-go"
)

func main() {
    client := va.NewClient(os.Getenv("VA_API_KEY"))

    stream, err := client.Models.Completions(context.Background(), &va.Request{
        Model:     "vareon-1",
        Messages:  []va.Message{{Role: "user", Content: "Analyze..."}},
        MaxTokens: 4096,
        Stream:    true,
        Context: va.Context{
            DVIProfile:    "user_12345",
            InstitutionID: "uni_trento_01",
        },
    })
    if err != nil { panic(err) }

    for chunk := range stream {
        fmt.Print(chunk.Delta.Content)
    }
}
Response text/event-stream
// Status 200 OK · text/event-stream

data: {
  "id": "resp_va_x92kq1",
  "model": "vareon-1",
  "object": "chat.completion.chunk",
  "choices": [{
    "delta": {
      "content": "Based on the DVI analysis..."
    },
    "finish_reason": null
  }],
  "usage": {
    "prompt_tokens": 142,
    "completion_tokens": 38
  },
  "vita_metadata": {
    "dvi_impact": "stability+2",
    "confidence": 0.94,
    "trace_id": "trace_va_8h2k1"
  }
}

data: {
  "choices": [{
    "delta": {
      "content": " the user is in stability mode."
    },
    "finish_reason": "stop"
  }]
}

data: [DONE]
06
Intelligence architecture
five layers · one mind
L1
L1
Input Processing
Text, documents, behavioral signals, and DVI profile data normalized into a single typed input.
Foundation
L2
L2
Context Assembly
Life-situation reconstruction, history continuity, and institutional context binding.
Context
L3
L3
DVI Reasoning
Stability · Load · Growth · Support — four-vector dimensional analysis routed to the weakest axis first.
Intelligence
L4
L4
Action Generation
Timed, specific, executable outputs — never generic advice. Every move ships with reasoning.
Execution
L5
L5
Consent Filter
Privacy-preserving output gating for institutional delivery — explicit, governed, revocable.
Governance
07
Model comparison
capability matrix
Capability Mitra 1.3G Mini Mitra 1.2 Vareon 1
Website generation ✓ Basic ✓ Advanced ✓ Premium
Life situation analysis ◐ Limited ✓ Full ✓ Institutional
DVI computation ◐ Individual ✓ Population
Investor mode ✓ Enhanced
Early risk detection
Institutional reporting
Multilingual output 14 langs 22 langs 42 langs
Context window 128K 200K 512K
Streaming output
API access ◐ Rate limited ✓ Standard ✓ Priority
› VitaAvanza Intelligence — Trento, Italy

Built for humans. Not just for tasks.

All models are developed and maintained by VitaAvanza S.r.l. We are committed to building AI that is transparent, purposeful, and human-centered.