🧠 AI Sovereign Hub 17 PROVIDERS · 0€ INFRA

Providers totaux
17
Cascade auto-fallback
Coût mensuel
0€
Free tiers + sovereign
Cascade tier active
● T1+T2
Port 4000 sovereign
Ollama local
9
Models LIVE S204
Last refresh
--:--
Auto every 60s
Brain Council
● READY
Parallel vote 5 IA

📊 Provider Usage Distribution (24h)

📈 Latency Comparison (avg ms)

🚀 Tier 1 — Primary Cascade (souverain · low latency · auto-fallback)

⚡ Tier 2 — Secondary (high quality · larger context)

💻 Ollama Local (S204) — fully sovereign · offline-capable

🧠 Brain Council Pattern — quand cascade fail, vote parallèle 5 IA

Cerebras
Qwen 235B
↘
Groq
Llama 3.3
↓
Consensus
Vote 3/5
↓
SambaNova
DeepSeek
↗
Ollama
local
+ Arena/AllIAHub via Blade cookie (5ème slot redondance)

📊 Cascade Health Live

Sovereign API port 4000checking...
Cerebras model activeqwen-3-235b-a22b-thinking-2507
Groq model activellama-3.3-70b-versatile
SambaNova modelMeta-Llama 3.3 70B
Ollama port 11434checking...
Auto-fallback chainCerebras → Groq → CF → Ollama
Avg cascade latency~880ms

🔑 API Keys Status

Cerebras API keySET
Groq API key⚠ rotate needed
SambaNova keySET
Gemini keyMISSING
CF Workers AISET
HuggingFace tokenSET
Anthropic keySET
OpenRouter / MoonshotMISSING (Kimi K2)
AI Sovereign Hub WEVAL · 17 providers cascade · Brain Council ready · 0€ infrastructure · ← WTP · All-IA Hub · Multi-Agent · Paperclip · DeerFlow