AI Engine & System Settings

Configure AI provider failover chains, API credentials, and Telegram error alerts.

1. Multi-Provider Priority Failover Chain (4 Stages)Reorder stages with ▲ / ▼ arrows to set priority order

The pipeline always queries Stage 1 first. Upon stage timeout or API error/rate-limit, it gracefully falls back down the chain (#1 → #2 → #3 → #4).

🥇Priority #1FIRST CHOICE (PRIMARY)

Provider:

Model Identifier:

Timeout (sec):

🥈Priority #2AUTO-FAILOVER #1 (SECONDARY)

Provider:

Model Identifier:

Timeout (sec):

🥉Priority #3AUTO-FAILOVER #2 (TERTIARY)

Provider:

Model Identifier:

Timeout (sec):

Enable Automatic Priority Failover across configured stages (#1 → #2 → #3)

2. API Keys Management (Encrypted Storage)

Save credentials for each provider. The grabber automatically attaches the correct key when switching between providers in the failover chain.

DeepSeek API Key

Google Gemini API Key

OpenRouter API Key

🛠️ Custom Endpoint Key & URL

Verify AI Models & Credentials

Test any priority stage with an instant live request to verify credentials and response latency.

📱 Telegram Error Monitoring & Live Alerts

Receive instant alerts on your phone whenever the automated news grabber hits critical errors or provider downtime.

🟢 Alerts Active

Telegram Bot Token (@WBS_status_bot)

Created via @BotFather in Telegram.

Telegram Chat ID / Group ID

Your personal chat ID or private channel ID.

Silence Alert Threshold (Hours)

Alert if 0 stories published in X hours (default: 5.0h, prevents night alerts).

Sends a verification ping to @WBS_status_bot.

🔍 Content Audit & Quality Control Engine

Configures the high-reasoning LLM utilized by the Audit & QA tab to fact-check published articles against source wires, check quote veracity, and detect AI-slop.

Audit AI Provider

Audit Model Identifier

Timeout (Seconds)

Recommended Reasoning Presets:

Verifies credentials, reasoning capacity, and latency for z-ai/glm-5.3-flash.