GPT-6.1 Sol Script Generation and Four New Voice Models Are Now Available
Script generation now uses GPT-6.1 Sol, and voice settings include ElevenLabs v4 Turbo/v4 and Gemini 3.8 Flash-Lite/Flash.
We have updated script and voice generation in SpeechSlide AI. The new models are available in the workflow from turning slides into speaker notes to creating a narrated video.
Script generation
Both single-slide and batch script generation now use OpenAI GPT-6.1 Sol. You can still attach a PDF for context and edit the generated scripts before creating audio.
Voice generation
We have added ElevenLabs v4 Turbo, ElevenLabs v4, Gemini 3.8 Flash-Lite, and Gemini 3.8 Flash. Choose a provider and model in a slide’s voice settings to generate narration from its script. For new slides, the default model is v4 Turbo when using ElevenLabs and 3.8 Flash-Lite when using Gemini. Model choices already saved on existing slides are preserved, and the earlier voice models remain available.
Generate or edit your script, then try the new models in voice settings. See the How to Use page for the basic workflow.