first audio thing that i want to try

README.md

@@ -78,6 +78,51 @@ docker login git.hiddenden.cafe
 docker push git.hiddenden.cafe/mats/cozy-den:latest
 ```
 
+## Blog Text-to-Speech
+
+Blog posts kunnen automatisch omgezet worden naar audio met `scripts/tts_generate.py`.
+
+### Setup
+
+1. Kopieer `.env.example` naar `.env` en vul je API key in
+2. Exporteer de variabelen:
+```bash
+export GOOGLE_API_KEY=jouw_sleutel_hier
+# of: export MISTRAL_API_KEY=... / export OPENAI_API_KEY=...
+```
+
+### Audio genereren
+
+```bash
+# Eén post — output: <slug>.mp3 in de huidige map
+python scripts/tts_generate.py src/content/blog/hello-world.md
+
+# Andere output directory
+python scripts/tts_generate.py src/content/blog/hello-world.md --output-dir public/audio
+
+# Andere provider
+TTS_PROVIDER=openai python scripts/tts_generate.py src/content/blog/hello-world.md
+```
+
+Beschikbare providers (via `TTS_PROVIDER`):
+
+| Provider | Env var | Prijs |
+|---|---|---|
+| `google` (standaard) | `GOOGLE_API_KEY` | 1M chars/maand gratis |
+| `mistral` | `MISTRAL_API_KEY` | ~$16/M chars |
+| `openai` | `OPENAI_API_KEY` | ~$15/M chars |
+
+### Audio tonen op een blogpost
+
+Voeg `audio: true` toe aan het frontmatter van de post. De `<audio>` player verschijnt automatisch en laadt van `https://audio.hiddenden.cafe/<slug>.mp3`.
+
+```yaml
+---
+title: "Mijn post"
+audio: true
+---
+```
+
 ## Customization
 
 The site is built to be easily customizable: