Doc · 06 AI text generation
AI text generation · ChatGPT, Claude, Mistral & Gemini
Descriptions, meta titles, meta descriptions and keywords at the push of a button - in every language. With ChatGPT, Claude, Mistral or Gemini, all usable in parallel with provider selection per generation.
Four providers, all directly integrated
| Provider | Models | Strengths |
|---|---|---|
| Anthropic Claude | Haiku 4.5 · Sonnet 4.6 · Opus 4.8 · Fable 5 | Longer, nuanced texts, brand style |
| OpenAI ChatGPT | GPT-5.5 · GPT-5.4 · GPT-5 mini · GPT-5 nano · GPT-4.1 mini | Fast, cost-effective generation |
| Mistral AI | Mistral Large 3 · Medium 3.5 · Small 4 | Strong European models, good value |
| Google Gemini | Gemini 3.1 Pro · 3.5 Flash · 3.1 Flash-Lite | Very fast flash models, large context windows |
When several are active, the generator modal lets you pick the provider per call - Claude for the demanding main description, a flash model for fast meta data.
Set up API key
- Create an API key at the respective provider: console.anthropic.com (Claude), platform.openai.com (ChatGPT), console.mistral.ai (Mistral) or Google AI Studio (Gemini)
- Enter in Config matrix → Translations in the respective provider section
- Click "Test connection" to verify (separately per provider)
- Pick model and tone default
- Click "Enable bulk" to activate in bulk-edit
Security: Keys are encrypted in the database and used server-side only. They never leave the browser. Own keys, no provider lock-in.
What can be generated
- Description (HTML) - formulated with paragraphs
- Meta title - max 60 chars, SEO-optimised
- Meta description - max 155 chars
- Keywords - comma-separated
Context chips · You control what the AI sees
- Name - always included
- Existing description as inspiration baseline
- Properties, custom fields, manufacturer, category
- Own keywords for SEO focus
Five style templates
- Professional, Casual-modern, Premium/Luxury, Technical, Emotional
AI in bulk-edit
Activatable per provider: all four AI providers can generate texts for many products in parallel - with concurrency control against rate limits.
Preview before applying
Generated texts appear in preview mode first. Review, optionally create snapshots, then apply. "Overwrite existing" is optional.
Costs and limits
- Token display: estimated tokens and costs in € per call
- Bulk limits: configurable upper bound
- Cost warning: threshold dialog at €5 by default
Troubleshooting
- Test API key under Config matrix → Translations → Test connection
- Check credit/billing status at the respective provider (Anthropic, OpenAI, Mistral, Google)
- Check server logs:
var/log/prod-*.log