VARKOS FLUX — SOVEREIGN ROUTER
◈ ONE ENDPOINT · SIX ENGINES · ZERO COMPLEXITY
Route smarter.
Build faster.
Send one request. FLUX automatically routes to the optimal intelligence engine based on what your request actually needs.
No model juggling. No fallback logic. No API key management.
// One endpoint. FLUX decides the rest.
POST https://varkos.vip/api/v1/router
x-api-key: vk_flux_your_key
{
"message": "Analyze the AI market in West Africa"
}
// FLUX auto-selected Engine α — deep strategy detected
{
"ok": true,
"reply": "West Africa's AI market...",
"provider": "engine-alpha",
"model": "depth-v4",
"routing": { "reason": "strategy", "latency_ms": 847 }
}
Selected: ENGINE α via FLUX
6
AI Engines
8
Routing rules
6x
Engine fallback
44
Languages
◈ MODELS
Multi-engine sovrano.
One intelligence layer.
FLUX knows each model's strengths. It picks the right one — automatically.
ENGINE α — DEPTH
Reasoning · Strategy · Code · Classified
Deep strategy & analysis
Complex code & debugging
Long-form reasoning
High-stakes decisions
ENGINE β — SPEED
Speed · Vision · Real-time · Tools
Fast conversational replies
Image & vision analysis
Real-time web search
Tool use & function calls
ENGINE γ — SCALE
Scale · Languages · Long documents · Volume
1M token context window
44 languages incl. African
Long document processing
Cost-optimized volume
ENGINE δ — REALTIME
Real-time · Streaming · Live · Low-latency
Ultra-low-latency replies
Live streaming responses
High-throughput bursts
Real-time agent ops
ENGINE ε — REASONING
Logic · Math · Step-by-step · Analysis
Advanced multi-step logic
Mathematical reasoning
Structured problem-solving
Verifiable chains of thought
ENGINE ζ — EU / SOVEREIGN
EU · Multilingual · Privacy · Compliance
EU-resident processing
Multilingual incl. African
Privacy & data sovereignty
Regulatory compliance
◈ ROUTING LOGIC
How FLUX decides
Every request is analyzed in milliseconds. No random. No round-robin. Pure intelligence.
| REQUEST TYPE | SIGNAL DETECTED | MODEL | REASON |
|---|---|---|---|
| Code / debugging | ```code, function, bug, error | Engine α | Superior reasoning |
| Strategy / analysis | strategy, market, competitor, SWOT | Engine α | Depth & quality |
| Document > 8k chars | Message length | Engine γ | 1M context window |
| Real-time / news | today, latest, current price | Engine β | Web search enabled |
| African languages | Wolof, Swahili, Hausa, Yoruba… | Engine γ | 44 language support |
| Vision / images | image, photo in request | Engine β | Native vision |
| Short / fast reply | Message < 120 chars | Engine β | Speed < 500ms |
| Classified / V•I•X | High-clearance plan | Engine α | No compromise |
+ Automatic 3-tier fallback. If a provider is down, FLUX switches silently.
◈ HOW IT WORKS
Up in 60 seconds
01
GET YOUR KEY
Sign up below. Free plan is instant — no card required. Your key starts with
vk_flux_02
SEND REQUEST
POST to
/api/v1/router with your message and API key. That's it. No model selection needed.03
FLUX ROUTES
FLUX analyzes your request, picks the optimal model, handles fallback. You get the best possible response — always.
04
BUILD & SCALE
One integration. Unlimited possibilities. Scale from 50 to 9,000 requests/day without changing a line of code.
◈ PRICING
Simple. Transparent.
Start free. Scale when you're ready. All paid plans include a 14-day trial.
FREE FOREVER
SPARK
€0
free · no card
- 50 requests/day
- Engine γ
- 300 tokens/request
- Basic routing
- provider_hint
- Routing details
- Priority support
SELF-SERVE
STREAM
€29/mo
14-day trial · card required
- 200 requests/day
- Engine γ + Engine β
- 1.200 tokens/request
- Smart routing
- provider_hint
- Routing details
- Priority support
SELF-SERVE
SURGE
€99/mo
14-day trial · card required
- 900 requests/day
- All 6 engines
- 1.500 tokens/request
- Full routing details
- provider_hint
- Conversation history
- Priority support
SCALE
VORTEX
€499/mo
14-day trial · card required
- 4.500 requests/day
- All engines
- 2.500 tokens/request
- Full routing details
- provider_hint
- Conversation history
- Priority support
ENTERPRISE
STORM
€999/mo
contract · wire or card
- 9.000 requests/day
- All engines + custom δ
- 4.000 tokens/request
- 99.9% SLA guaranteed
- Custom routing rules
- White-label · your brand
- Dedicated line to KOS
◈ 14-day free trial on STREAM and SURGE — charged only after trial ends
◈ CODE
Start in 60 seconds
JavaScript
Python
cURL
Advanced
// JavaScript / Node.js const res = await fetch('https://varkos.vip/api/v1/router', { method: 'POST', headers: { 'Content-Type': 'application/json', 'x-api-key': 'vk_flux_your_key' }, body: JSON.stringify({ message: 'Analyze fintech trends in Nigeria 2026' // Optional: provider_hint: 'alpha' | 'beta' | 'gamma' }) }) const { reply, provider, routing } = await res.json() console.log(reply) // AI response console.log(provider) // 'alpha' | 'beta' | 'gamma' console.log(routing.reason) // 'strategy' → why FLUX was chosen
# Python import requests r = requests.post( 'https://varkos.vip/api/v1/router', headers={ 'x-api-key': 'vk_flux_your_key', 'Content-Type': 'application/json' }, json={ 'message': 'Write a Python function to clean CSV data', 'system': 'You are a senior data engineer', 'options': {'max_tokens': 2000} } ) data = r.json() print(data['reply']) print(f"Provider: {data['provider']} · {data['usage']['latency_ms']}ms")
# cURL curl -X POST https://varkos.vip/api/v1/router \ -H "x-api-key: vk_flux_your_key" \ -H "Content-Type: application/json" \ -d '{"message": "Latest AI news today"}' # Response — FLUX chose Engine β (realtime signal detected) { "ok": true, "reply": "Today in AI...", "provider": "engine-beta", "model": "speed-v4", "usage": { "latency_ms": 412, "remaining": 4998 } }
// Advanced — history + system + force provider const result = await fetch('https://varkos.vip/api/v1/router', { method: 'POST', headers: { 'x-api-key': 'vk_flux_surge_key', 'Content-Type': 'application/json' }, body: JSON.stringify({ message: 'Continue — what are the main risks?', system: 'You are a senior investment analyst', history: [ { role: 'user', content: 'Analyze African fintech market' }, { role: 'assistant', content: 'The African fintech market...' } ], // Force a specific provider (SURGE+ only) provider_hint: 'alpha', options: { max_tokens: 2000 } }) }) const { reply, routing } = await result.json() // routing.classification.is_strategy === true // routing.selected_provider === 'alpha' // routing.latency_ms === 1240
◈ FAQ
Questions answered
Do I need API keys for multiple AI providers separately?
No. That's the whole point of FLUX. You have one VARKOS FLUX key and VARKOS handles all provider connections internally. You never manage provider keys or accounts.
Can I force a specific model?
Yes — on STREAM and above you can pass
provider_hint: "alpha" (or "beta" or "gamma") to override the auto-routing. Without it, FLUX always picks the best model for your request.What happens if a provider is down?
FLUX automatically falls back to the next best engine — silently. If FLUX is unavailable, it tries Engine β, then Engine γ. You always get a response. This 3-tier fallback is included on all plans.
Is FLUX different from the VARKOS API?
The VARKOS API gives you access to 200+ specialized agents. FLUX is the intelligent routing layer — a single endpoint that routes between FLUX, Engine β and Engine γ. They're complementary. Many clients use both.
Is there a white-label option?
Yes — on the STORM contract plan. You can integrate FLUX under your own brand, with your own domain and API endpoint. Contact us at hello@varkos.vip for details.