AI / Google Antigravity Gemini Fundamentals Interview Questions
What are the Gemini API's multimodal input capabilities?
Gemini models are natively multimodal - they can accept and reason across text, images, video, audio, and documents (PDFs) in a single request. This is a fundamental design characteristic, not a bolt-on feature.
| Modality | Supported formats | Notes |
|---|---|---|
| Text | Plain text, markdown | Always supported |
| Images | JPEG, PNG, WEBP, HEIC, HEIF | Up to ~3,600 per request (model-dependent) |
| Video | MP4, MOV, AVI, WMOV, MPEG, WebM, FLV | Up to 1 hour of video in context |
| Audio | MP3, WAV, FLAC, AAC, OGG, OPUS, WebM | Transcription, analysis, audio Q&A |
| application/pdf | Documents with text + embedded images | |
| Structured data | application/json, text/csv | Data files for analysis |
from google import genai from pathlib import Path client = genai.Client() # Upload a file via the Files API (for large files) my_video = client.files.upload(file=Path("presentation.mp4")) # Multimodal request: video + text question interaction = client.interactions.create( model="gemini-3.5-flash", input=[ {"type": "text", "text": "Summarise the key points from this video."}, {"type": "video", "uri": my_video.uri, "mime_type": "video/mp4"} ] ) print(interaction.output_text) # Image inline (base64 for small images, Files API for large): import base64 image_data = base64.b64encode(Path("diagram.png").read_bytes()).decode() interaction = client.interactions.create( model="gemini-3.5-flash", input=[ {"type": "text", "text": "Explain this architecture diagram."}, {"type": "image", "data": image_data, "mime_type": "image/png"} ] )
What is the recommended way to include large video files in a Gemini API request?
Which of the following is NOT currently a supported input modality for Gemini models?
More Related questions...
