AI assistant

Effess includes an AI assistant powered by your own API key. No middleware, no proxying, no per-message billing from us — you pay your AI provider directly.

AI assistant panel

Supported providers

• OpenAI — GPT-4o, GPT-4o-mini, GPT-3.5
• Anthropic — Claude Opus 4.7, Claude Sonnet 4.6, Claude Haiku 4.5
• Groq — Llama 3.1, Mixtral 8x7B (free tier — very fast)
• Google — Gemini 1.5 Pro / Flash

Set up your key

1. Tools → Settings → AI.
2. Paste your API key.
3. Pick a default model.
4. Save.

The key is stored encrypted in chrome.storage.local. It is never sent to any Effess server.

What the AI can do

In any chat, hover over the WhatsApp composer to reveal the Effess AI bar with these one-click actions:

• Translate to… — pick a target language; the AI translates your draft.
• Smart reply — generates 3 candidate replies based on the last 5 messages.
• Summarise chat — TL;DR the last 20-50 messages.
• Tone polish — make your draft more formal / casual / friendly / concise.
• Continue writing — finish your half-typed message.

Custom prompts

Tools → Templates → + New AI prompt. Define a system prompt and call it by slash (/qualify, /objection, /upsell). The AI runs your prompt with the chat history as context.

Cost & limits

You pay the API provider directly. Effess never sees the message bodies or AI responses — they go straight from your browser to the AI API.

Approximate costs (varies by model):

• Translate 1 message: $0.0001 - $0.001
• Summarise 50 messages: $0.005 - $0.02
• Smart reply (3 candidates): $0.001 - $0.005

Free Groq tier gets you several thousand requests/day at zero cost.