Gemini 3.8 TTS Hands-On: Emotion and Dialect Markers Enable Free Arabic Voiceover
aziz4ai · x · 2026-09-24
A hands-on look at the new Gemini 3.8 TTS: strong Saudi Arabic dialect support, free voiceover generation, plus voice cloning and voice design. The key technique — wrap emotions in parentheses and dialects in <> tags before the script — lets users quickly produce natural-sounding Arabic narration, though cloning a specific voice requires a paid tier.
More from Multimodal
- DeepMind ships Gemini 3.8 Flash TTS with native multi-speaker overlap and laughter — kastnerkyle · 2026-09-24
- Treating AI video as a production pipeline: building a reusable cinematic producer agent in CREAO — FellMentKE · 2026-09-24
- Open-source ComfyUI node restores lost template filters for Local, Partner and Credit workflows — linus74RN · 2026-09-24
- Gemini 3.8 Flash TTS demo nails the Osaka auntie accent — heiga_zen · 2026-09-24
- First/last-frame video workflow with GPT Image + LTX, author wants to move to local Qwen Image 2.1 — ART-ficial-Ignorance · 2026-09-24
- ElevenLabs MCP adds voice, music, image and video generation to your AI assistant — aftahi_ai · 2026-09-24