Gemini 1.5 Pro — Google's Multimodal Model Powering HYR Studio's Video Layer
The only AI model that watches your video — not just reads the transcript — to build richer marketing assets.
Gemini 1.5 Pro from Google DeepMind is a natively multimodal model capable of processing video, audio, images, and text in a single context window of up to 1 million tokens. HYR Studio uses Gemini for tasks where the source is a video or audio file and rich visual or audio understanding is needed.
Provider
Google DeepMind
How HYR Studio uses Gemini 1.5 Pro
HYR Studio routes Gemini 1.5 Pro when the source material is a video or podcast recording. Gemini processes the full video — not just the transcript — picking up visual context, speaker tone, and on-screen content that text-only models would miss. This enables richer social clips descriptions, more accurate show notes, and better visual asset recommendations.
Capabilities
- 1M token context window (longest in production)
- Native video and audio understanding
- Frame-by-frame visual analysis
- Multi-speaker audio diarization
- Cross-modal reasoning (text + image + audio)
- Real-time streaming output
Output Types Powered
- Video clip summaries
- Podcast episode notes
- Visual asset briefs
- Multi-platform captions
- Speaker highlight reels
- Transcript-aligned timestamps