Step 1: Choose prompt type
Select chat, freeform, or structured format based on your use case
Step 2: Configure model and parameters
Pick a Gemini model, set temperature, top K/P, safety, and system instructions
Step 3: Add multimodal inputs
Upload images, audio, video, PDFs, or code files alongside your text prompt
Step 4: Iterate and refine
Test variations, adjust parameters, and compare responses across models
Step 5: Export production code
Click "Get code" to generate API call snippets in Python, JavaScript, Go, Java, C#, or REST
Supported input modalities
Text
Prompts, system instructions, few shot examples, and structured schemas.
Images
PNG, JPG, GIF, WebP. Vision understanding, OCR, diagram analysis.
Audio
MP3, WAV, FLAC. Transcription, analysis, and audio understanding.
Video
MP4, MOV. Scene understanding, content analysis, and summarization.
Documents
PDFs up to 1,000 pages with full multimodal understanding.
Code
Source files in any language. Analysis, review, and generation.