ai-multimodal

Audio transcription with Gemini API converts speech to text and extracts key information from audio files up to 9.5 hours long. The Skill handles transcription, speaker identification, and content analysis in a single workflow, returning structured text or JSON output for downstream processing.

Yes, Gemini's vision capabilities extract objects, text (OCR), and visual insights from images and video frames. This Skill automates frame extraction, analysis, and synthesis of visual data across batches, supporting sentiment detection and object segmentation in a single workflow.

This Skill converts and optimizes various document types for Gemini API consumption, handling format conversion and media optimization. Outputs are generated in text, JSON, CSV, or Markdown formats, enabling direct ingestion into AI pipelines without manual preprocessing.

Multimodal analysis with Gemini processes audio transcription, video frame extraction, and image analysis simultaneously to synthesize cross-modal insights. This Skill enforces media constraints (9.5 hours audio, 6 hours video) and returns integrated reports combining all modalities.

Gemini API enforces media length constraints: audio is limited to 9.5 hours and video to 6 hours. This Skill respects these boundaries during batch processing; exceeding limits requires splitting content before submission to the API.

Yes, this Skill supports both File API uploads and inline media inputs for flexible integration. File API suits large batches and persistent storage; inline inputs work for real-time requests, enabling hybrid workflows tailored to your processing scale.